Competition in Electricity Markets - Regulation Body of

I N T E R N AT I O N A LE N E RGY AG E N CY

ENERGYMARKETREFORM

COMPETITIONIN ELECTRICITY

MARKETS

FTCompetition 15/01/01 11:52

I N T E R N AT I O N A LE N E RGY AG E N CY

ENERGYMARKETREFORM

COMPETITIONIN ELECTRICITY

MARKETS

gardeCompet 15/01/01 17:29 Page 1

INTERNATIONALENERGY AGENCY

9, rue de la Fédération,75739 Paris Cedex 15, France

The International Energy Agency (IEA) is anautonomous body which was established inNovember 1974 within the framework of theOrganisation for Economic Co-operation andDevelopment (OECD) to implement an interna-tional energy programme.It carries out a comprehensive programme ofenergy co-operation among twenty-five* of theOECD’s thirty Member countries.The basic aims of the IEA are:• To maintain and improve systems for coping

with oil supply disruptions ;• To promote rational energy policies in a glo-

bal context through co-operative relationswith non-member countries, industry andinternational organisations ;

• To operate a permanent information systemon the international oil market ;

• To improve the world’s energy supply anddemand structure by developing alternativeenergy sources and increasing the efficiencyof energy use ;

• To assist in the integration of environmentaland energy policies.

*IEA Member countries : Australia, Austria,Belgium, Canada, the Czech Republic,Denmark, Finland, France, Germany, Greece,Hungary, Ireland, Italy, Japan, Luxembourg,the Netherlands, New Zealand, Norway,Portugal, Spain, Sweden, Switzerland, Turkey,the United Kingdom, the United States. TheEuropean Commission also takes part in thework of the IEA.

ORGANISATION FORECONOMIC CO-OPERATION

AND DEVELOPMENT

Pursuant to Article 1 of the Convention signedin Paris on 14th December 1960, and whichcame into force on 30th September 1961, theOrganisation for Economic Co-operation andDevelopment (OECD) shall promote policiesdesigned :• To achieve the highest sustainable economic

growth and employment and a rising stan-dard of living in Member countries, whilemaintaining financial stability, and thus tocontribute to the development of the worldeconomy ;

• To contribute to sound economic expansionin Member as well as non-member countriesin the process of economic development ; and

• To contribute to the expansion of worldtrade on a multilateral, non-discriminatorybasis in accordance with international obli-gations.

The original Member countries of the OECDare Austria, Belgium, Canada, Denmark,France, Germany, Greece, Iceland, Ireland,Italy, Luxembourg, the Netherlands, Norway,Portugal, Spain, Sweden, Switzerland, Turkey,the United Kingdom and the United States. The following countries became Memberssubsequently through accession at the datesindicated hereafter : Japan (28th April 1964),Finland (28th January 1969), Australia (7thJune 1971), New Zealand (29th May 1973),Mexico (18th May 1994), the Czech Republic(21st December 1995), Hungary (7th May1996), Poland (22nd November 1996), theRepublic of Korea (12th December 1996) andSlovakia (28th September 2000). TheCommission of the European Communitiestakes part in the work of the OECD (Article 13of the OECD Convention).

© OECD/IEA, 2001

Applications for permission to reproduce or translate all or part of this publicationshould be made to : Head of Publications Service, OECD

2, rue André-Pascal, 75775 Paris cedex 16, France.

FOREWORD

Most OECD countries, and many countries outside the OECD, areintroducing competition into their electricity markets.This is a challengingprocess. Promoting effective and sustainable competition requires actionon a number of related issues and an overhaul of traditional marketstructures and regulatory frameworks.

This book considers the experience of OECD countries in the reform oftheir electricity markets. Its main focus is the introduction of competition.Reform also has significant implications for other key policy issues suchas security and the environment, but it is not the main purpose here toanalyse these.

Reform experience so far highlights a converging trend towardsintroducing consumer choice of electricity supplier as a fundamentalpillar of effective reform.This means stimulating competition not only ingeneration, but also in electricity trading and supply as a service toconsumers. At the same time, the need to regulate transmission systemseffectively is becoming evident, so as to ensure a level playing field formarket participants and sustain efficient investment in transmission.Many countries are also engaged in a major overhaul of regulatoryinstitutions to oversee the new markets.

As well as providing a strategic and comprehensive overview of thereform process, this book contains a detailed analysis of key elements,including unbundling the network from potentially competitive activities,the evolution of competitive electricity spot markets, the regulation andpricing of transmission, and the role of system operators.These issues arein full evolution and approaches continue to be improved and refined.The book will therefore be of interest to energy policy makers, legislatorsand electricity specialists.

It is important for reforming countries to keep track of developmentsand best practices.The IEA monitors, and will continue to monitor, theevolution of electricity market reform. Developments are reportedregularly in the IEA’s Annual Energy Policy Review Book.

3

competition in electricity markets

This book has benefited enormously from discussions at meetings of theIEA’s Electricity Regulatory Forum on such issues as electricity tradingand transmission pricing. I would like to thank the IEA member countryparticipants of these meetings. The main author of the book is CarlosOcaña. Caroline Varley directed the work and provided editorial oversight.

This book is published under my authority as Executive Director of theInternational Energy Agency.

Robert PriddleExecutive Director

4

competition in electricity markets ===

5


TABLE OF CONTENTS

INTRODUCTION AND EXECUTIVE SUMMARY 9

A Common Approach to Reform is Developing 9A Review of Reforms to Date 10Key Issues: Unbundling 12Key Issues: Empowering the End User 13Key Issues: Meeting Security of Supply, Environmental and Social Goals 13Key Issues: Reforming Regulatory Institutions 14Designing Markets and Regulation 14Outlook 15

THE BACKGROUND TO REFORM 17

Defining Electricity 17What are the Benefits of Reform? 22What are the Costs of Reform? 25What is Driving Reform? 26

A REVIEW OF ESI REFORMS IN OECDCOUNTRIES 29

A Brief History 29Early Reforms 30Recent Reforms 36Performance 48

WHAT MODEL FOR THE ESI? 55

Retail Competition 55Alternative Approaches 58Retail Competition Issues 59An Assessment 62AppendixTransition Issues: Stranded Costs 65

4

3

2

1

6


UNBUNDLING 69

Why Unbundle? 69Vertical Separation of Generation and Transmission 70Vertical Separation of Generation and Distribution 76Vertical Separation of Distribution and End-user Supply 77

MARKETS 79

“Spot” Electricity Markets 79Power Exchanges: Review of the International Experience 82Trading Outside the Pool 91Capacity Mechanisms 93Financial Markets 97

NETWORKS 101

Means and Ends of Network Regulation 101Short-term Pricing Issues: Managing Congestion 104Setting Price Levels: Cost of Service Versus IncentiveRegulation 110The Institutional Framework: The Role of System Operators in Investment 113AppendixAllocating Costs to Users 114

INSTITUTIONS AND THE POLICYFRAMEWORK 117

The Need for Regulatory Institutions to Adapt 117Regulatory Independence 118A Review of Institutional Approaches 119The Important Role of Competition Policy 120Developing a Framework for International Trade 123

8

7

6

5

7


The Relationship with Natural Gas 125The Broader Policy Framework: Security of Supply 125The Broader Policy Framework: the Environment 128The Broader Policy Framework: Social and Regional Objectives 135

CONCLUSIONS AND OUTLOOK 137

Consumer Choice is the Key to Reform 137Challenges 137Further Change Can be Expected 140

REFERENCES 143

GLOSSARY 151

LIST OF TABLES

1. Functional Structure of the ESI 18

2. Key Characteristics of Some Early Reformers 31

3. Horizontal Concentration 34

4. Market Opening under the EU Electricity Directive 37

5. Network Access in EU Countries 40

6. European Electricity Market Reform 41

7. Wholesale Average Prices (US$ per MWh) 53

8. Cost of Metering Equipment (US$) 61

9

8


09. Organisation of Transmission-related Activities inCompetitive Electricity Systems 72

10. Organisation of Electricity Exchanges in the OECD Area 83

11. International Comparison of Transmission Pricing 103

12. Pricing of Network Services 104

LIST OF FIGURES

1. Planned Electricity Market Opening in EU Countries 38

2. Status of State Electricity Industry Restructuring as of May 2000 47

3. Market Share of Two Largest Generators in SelectedOECD Markets (1998) 49

4. Monopoly vs. Retail Competition 56

5. Retail Competition: How does it Work? 57

6. Other Approaches 57

LIST OF BOXES

1. Main Lines of Reform in the US 45

2. Competition for Small Consumers : the Case of California 62

3. How an Ideal Spot Electricity Market Would Work 80

INTRODUCTION AND EXECUTIVE SUMMARY

Virtually all OECD countries have decided to open up theirelectricity markets, at least to their big industrial users. In manycountries electricity markets will be open to all users, includinghouseholds. This is already the case in Finland, Germany, NewZealand,Norway, Sweden, England and Wales in the UK, and severalstates in the US and Australia. By the year 2006, more than 500 million people (and all large industrial users) in the OECD areawill be entitled to choose their electricity supplier. This accountsfor nearly 50% of the population of OECD countries.

This book analyses the development of choice and competition inthe Electricity Supply Industry (ESI). Drawing on a review of theinternational experience, it describes the main approaches that arebeing developed, discusses the key issues in the effective reform ofelectricity markets and provides an assessment of the emergingapproach to reform.The book is written from the perspective ofregulators and policy makers. It seeks to answer the question: whatis an effective regulatory framework for competition in electricitymarkets?

A common Approach to Reform is Developing

In principle, many different approaches to ESI reform are possibledepending on which activities are liberalised (e.g. generation), howthe non-liberalised activities (e.g. transmission) are regulated andwhich agents are allowed to participate in the different markets.However, in practice, there is increasing convergence in theapproach to reforming electricity markets in the OECD area.Recent ESI reforms often share these four elements:

■ Rapid introduction of full consumer choice;

9

introduction and executive summary1

■ An obligation to provide non-discriminatory Third-Party Access(TPA) to the transmission and distribution networks;

■ Unbundling of transmission; and

■ Liberalisation of electricity trade so that electricity can betraded both through organised power exchanges and on abilateral basis

This combination of full market opening, unbundling oftransmission activities, regulated access to the network andliberalisation of electricity trade is known as “retail competition”.Under retail competition, transactions among generators, endusers and a number of possible intermediaries, such as retailers,power exchanges and brokers take place freely (within the“physical” constraints imposed by the network). Thus, on thedemand side, end users are free to choose their supplier and tonegotiate their contracts; on the supply side, generators can selltheir electricity to any other market players. Retail competition hasinspired reforms in Finland, Norway, Spain, Sweden, UK, US and,with some variations, in Australia, Denmark, Germany and NewZealand.

A Review of Reforms to Date

Electricity reforms are expected to increase the efficiency of theESI. However, the most significant impacts of reform are onlyexpected to emerge in the long-term as a result of betterinvestment decisions. In most countries competition started veryrecently, so it is still too early to evaluate the performance — interms of costs, prices, and global social benefits — of the newelectricity markets.

In a short-term perspective, reforms have generally delivered theirexpected benefits. Large productivity increases have been reportedin a number of countries, largely linked to the corporatisation andprivatisation of the utilities. Final electricity prices have decreasedor remained stable in all countries examined and wholesale

10

introduction and executive summary === 1

electricity prices have been low (e.g. relative to the cost of newgeneration) whenever market power has not been an issue.However, as a result of lower costs, electricity prices have alsodecreased in many OECD countries that did not introducereforms until very recently, making it difficult to quantify the actual impact of competition on prices.

Market structure is a key determinant of prices in the newelectricity markets. High concentration of generation assets hasresulted in weak competition in the wholesale market (e.g. earlyUK and Spain), while more intense competition has taken place inless concentrated markets (e.g. the Nordic electricity market andGermany). The implication for policy is that reforms need to payattention to structural competition policies such as divestituresand/or the opening of national markets to international (orinterstate) trade and competition. Regulatory reform alone is notenough for competition to emerge.

A key issue for the progress of reforms is the distribution of thecosts and benefits of reform among end users, investors and otherparties such as tax payers and ESI employees. The distribution ofcosts and benefits has a large impact on the social and politicalacceptability of reforms, in addition to its impact on efficiency. Forinstance, in a number of countries cost and price reductions alongthe supply chain have not been fully reflected in end-user prices.This has raised concerns about the fairness of reforms andencouraged the introduction of retail competition and othermeasures to lower electricity prices. Also, the treatment ofstranded costs and, where applicable, the form of privatisation, havehad a significant impact on the distribution of the benefits ofreform.

The transition from the old to the new regulatory regimes poses amajor challenge for policy makers. Regulatory uncertainty duringor immediately before the transition may have a negative impact oninvestment as potential investors delay their decisions until thenew framework is defined.Also, the reliability of electricity supplymay decrease if the rules and responsibilities of the new actors are

11


not clearly and consistently defined. Experience shows thatreliability problems have been at most sporadic in reformedmarkets and are only partly attributable to transition issues.However, experience also shows that these problems, wheneverthey have materialised, have a large negative impact on the publicperception of reforms. Thus, governments have a key role duringthe transition in ensuring that the appropriate safeguards to sustainreliability are in place and in minimising regulatory risk.

Key Issues: Unbundling

In order for competition to develop in electricity markets,monopolistic activities such as the operation of the transmissionnetwork need to be effectively separated from the potentiallycompetitive activities (e.g. generation). The main objective ofunbundling is to avoid discrimination in the competitive segmentsof the ESI. Thus, some degree of separation is needed betweentransmission and generation, distribution and generation, anddistribution and end user supply.

Ownership separation or “divestiture” — requiring differentowners for different activities — has the greatest potential toeliminate discrimination, because it eliminates the incentive todiscriminate. Functional separation and accounting separation havea limited potential to prevent discrimination because the incentiveto discriminate and some of the ability to discriminate remain. Ifapplied, these forms of separation require significant regulatoryoversight and vigorous enforcement of competition law.Operational separation of transmission — separating systemoperation from the ownership of transmission assets — mayprovide a workable alternative to divestiture when transmissionownership is fragmented among several parties. However, theestablishment of effective independent system operators requiresthe development of complex and still largely untested governancestructures.

12


Key Issues: Empowering the End User

Despite its relatively small weight in the supply chain, end usersupply has a disproportionately large importance in gettingcompetition to work for the benefit of consumers. The ability ofend users to choose a supplier creates a fundamental pressure onall the players along the supply chain that is virtually impossible toreplicate by regulation.The value of consumer choice in discipliningmarket players is that it provides consumers with an effectivebargaining tool; the tool may be effective even if the options that itprovides (e.g. switching supplier) are not systematically used. Evenif this pressure is not directly observable (many consumers maychoose to remain with the same supplier), indirect effects on pricestructures, price levels, product diversity and service conditions arepotentially significant.

Introducing competition in end user supply requires unbundling itfrom distribution, a critical mass of suppliers to enable genuinechoice, and the development of an appropriate technicalframework related to metering and billing. In addition, there aretechniques, such as load profiling, that significantly reduce the costof introducing competition for small consumers.

Key Issues: Meeting Security of Supply,Environmental and Social Goals

In addition to economic efficiency, energy policy aims to meetother objectives such as security of supply, environmentalprotection and social goals. The old command and controlmechanisms traditionally applied to pursue these goals aregenerally neither the best approach nor feasible in the newcontext. In a competitive ESI, policies have to be implemented withcompetitively neutral instruments that do not discriminate amongmarket players and minimise market distortions. Implementingsuitable instruments has proved to be a difficult task.

13


Key Issues: Reforming RegulatoryInstitutions

Regulatory institutions need to adapt to meet the new challengesposed by reforms. In particular, regulators need to be independentfrom the regulated. Otherwise, conflicts of interest inevitably arise.Independence from government helps to ensure stability ofregulatory policies, to avoid the use of electricity policy to achievegeneral policy goals and, where government is also the owner ofutilities, to ensure impartial treatment of market players. However,ensuring the accountability of independent regulatory agencies is adifficult issue, and the choice of approach may depend on specificcountry features — for example, the role of the courts. In addition,the active role of competition authorities is needed as competitiondevelops.Two areas in which competition plays an important roleare merger control and the elimination of subsidies.

Designing Markets and Regulation

Together with the “big” strategic issues, the development ofelectricity markets necessitates addressing other more technicalissues concerning the regulation of transmission and distribution,and the organisation of the market. Chapters 6 and 7 provide anon-technical introduction to these issues along with a brief surveyof current approaches in OECD countries.

A key aspect of the regulation of the network is the pricing oftransmission. Nodal pricing provides incentives for an efficient useof generation and transmission assets. Experience shows that nodalpricing is workable, and its use may be expected to increaseprogressively. Postage stamp pricing does not generally provideadequate incentives for efficiency. However, inefficiencies may besmall in systems with a strong grid and large reserve generationmargins and postage stamp pricing has the advantage of beingrelatively transparent and easy to implement.

14


Two issues have polarised the debate on the organisation ofwholesale electricity markets. First, there is the question ofwhether mandatory or voluntary electricity pools should bepreferred.There is a growing consensus that competitive bilateralelectricity trading is an essential part of an efficient modernelectricity market. Voluntary pools or power exchanges areincreasingly dominating the scene, and mandatory pools arereceding. Although pricing and scheduling rules show significantvariation across systems, these differences do not seem to have asignificant impact on the functioning of electricity pools or powerexchanges. Second, some electricity markets have instituted so-called capacity payments to provide additional incentives forinvestment in generation while, in other markets, generators arepaid only for the energy actually provided. This study concludesthat, under most circumstances, market incentives to invest aresufficient to ensure adequate investment and that capacitypayments are not generally needed.

Outlook

Electricity market reform is a moving target. One key developmentis the market itself. The ESI is rapidly expanding its boundaries.Generation is increasingly integrated with gas and oil companies.At the same time, distribution and end user supply activities innetwork industries, such as gas, electricity, telecommunications orwater, are likely to become more integrated across industries.Thegeographical boundaries of the ESI, once coincident with nationalor state boundaries, are also changing. More and more, electricitysystems are becoming integrated within regional markets.Liberalisation is also opening the way for significant directinvestment by foreign companies in national markets. Theimplication for policy of this expanding industry boundary is thatregulation will have to be managed at a multi-sectoral andmultinational level.This evolution towards greater dependence oncompetition law and common rules for international trade is a

15


process that has already occurred in many other industries openedto competition.

The emergence of non-centralised dispatch is also changing theESI. Autoproduction and distributed generation are rapidlygrowing in many countries due to the development of efficientsmall scale generation. Environmentally these developments arevery positive to the extent that they make use of renewablesources of energy. Distributed generation and autoproduction areboth substitutes for electricity transportation services.This has theeffect of weakening the effectiveness of natural monopolyregulation of the network. In due course, if new generationtechnologies become a profitable substitute for transportation,transportation will cease to be a natural monopoly. Such adevelopment would have a crucial impact on the future economics,market structure and regulation of the ESI.

16


THE BACKGROUND TO REFORM

Defining Electricity

It is helpful to start by considering the different functions that makeup electricity delivered to the consumer. Some of these functionscan be supplied competitively, whereas others are more difficult toliberalise. Electricity delivered to the consumer is made up ofenergy (a non-storable commodity), and transportation (a service)which includes transmission, distribution and system operation.Ancillary services supply some special forms of energy needed bythe system operator to secure the short-term balance, securityand reliability of the system. Electricity is supplied to end users bybundling it with end user services (e.g. billing). Finally, there areother associated services (e.g. construction and maintenance).Thefunctional structure of the ESI is summarised in Table 1.

■ The Commodity: Energy

The commodity component of electricity is similar to many othercommodities although it has some special features. Electricitydemand fluctuates in the various time horizons (in a day, a year, orin the business cycle) both randomly and non-randomly. In additionelectricity, at present, cannot be economically stored. This meansthat:

■ Generation (and transmission) capacity needed to cope withpeak demand is partly unused in periods of lower demand;

■ Reserve capacity may be required to cope with randomdemand fluctuations or generation shortfalls; and

■ A diversified portfolio of electricity generating technologies isneeded to provide the different loads of electricity at least cost.

17

the background to reform2

18

the background to reform === 2

Function Key Economic Characteristics Implications

Generation • Limited scale economies Potentially competitiveat plant level

• Co-ordination economies at system level

• Complementarity with transmission

Transmission • Network externalities • Investment incentives• In general not a natural need special

monopoly attention

• Large sunk costs • One grid but possiblyseveral owners

Distribution • Often a natural monopoly No competition

• Large sunk costs

System Operation • Monopoly (due to technical No competitionconstraints)

End user Supply • Limited scale economies Potentially competitive

• No special features

Related Services:

• Power Exchanges No special features Potentially competitive

• Financial Contracts

• Construction and maintenance of assets

Table 1

Functional Structure of the ESI

Since the costs of electricity production from generating plantsvary, electricity generation is characterised by a merit order ofgenerating plants.There are related economies of co-ordination atthe system level. In varying degrees, generating technologies arealso characterised by their relatively high capital intensity, andtechnical and economic longevity, including long lead andconstruction times. However, some recent innovations aredramatically lowering capital intensity and shortening lead andconstruction times for some technologies (e.g. combined cycle gasturbines). Overall, economies of scale in generation do not seemto be significant at the plant level.

■ Transportation: Transmission and Distribution

It is customary to distinguish between two types of transportation:transmission is transportation at very high voltage levels anddistribution is transportation at lower voltage levels.Transmissionrefers to transportation over an interconnected network, which isshared by all end users, whereas distribution refers totransportation from the interconnected network to a specificgroup of end users; a transmission line thus provides security ofsupply to all end users while a distribution line benefits only some.

Distribution lines are often considered a natural monopoly sinceduplication of distribution lines would be inefficient due to thelarge fixed costs of the investment. There are exceptions such asbuildings and factories that have two connections to thedistribution grid to ensure security of supply.

The transmission network also has some special characteristics.First, there are so-called network externalities, i.e. investmentsbenefit all interconnected parties by increasing reliability andsecurity and reducing the cost of generation. Networkexternalities may result in the additional value of investments ingrid augmentation being reduced by successive investments. Thismay discourage investment. Second, there are system-wideeconomies of scale as in the case of distribution.

19


However, transmission lines within the grid are not, in general,natural monopolies.Two transmission lines may run more or lessin parallel and still be economical; and two nodes within aninterconnected grid are often connected through several paths as a means to increase reliability. Thus, transmission services canbe provided by different owners within a single interconnectednetwork.

■ System Operation

System operation refers to the co-ordination of transportationservices to ensure that the system is constantly in state of staticelectrical equilibrium. In particular, equilibrium requires that powersupplied equals power demanded at each node of the network.This state is achieved by controlling inflows and outflows of energyover the network and by procuring the complementary ancillaryservices necessary to maintain the technical reliability of the grid.The scope of system operation changes with the regulatoryframework. Decisions made at the time of delivery are alwayscontrolled by the system operator, while decisions made sometime in advance of actual delivery can be made either by the system operator or by market participants. The development of information technology is quickly shortening the period overwhich decisions have to be made by the system operator. Forinstance, the Australian electricity market is open up to fiveminutes before delivery.

Regardless of the market framework (monopoly or competitive),system operation always remains a monopoly. Interconnection andits associated benefits of increased reliability and lower costs areonly possible under a centralised system operation. However, thisintrinsically monopolistic function can be unbundled fromtransmission ownership, in which case an independent systemoperator is in charge of system operation.

20


■ End-user Supply and Services

End-user supply refers to the delivery of electricity to end users. Itincludes the procurement of energy and transportation servicesand the metering and billing of consumption. End-user supply wastraditionally bundled with distribution but can be performedseparately.There is an increasing number of “value-added services”linked to end-user supply such as supplying differentiatedelectricity (e.g. green energy), packaging electricity (e.g. with otherutility services such as gas) and supplying differentiated reliabilityand quality (e.g. interruptible supply). End-user services arepotentially competitive.

Suppliers to end-users thus perform two functions. First, they actas brokers who buy and sell energy and try to make a profit fromassuming the risk of price volatility and from adjusting prices toconsumption patterns1. Second, suppliers may provide the servicesto end-users mentioned above.

■ Related Services

Electricity supply involves other activities. Construction andmaintenance services provide and maintain the generating plantsand grid assets needed to supply electricity. There are also manynew financial services, such as those offered by a growing numberof power exchanges, that facilitate trade in electricity. Financialinstruments (e.g. electricity futures) are also being developed toprovide for a better management of risks. These services arepotentially competitive subject to the same type of regulation thatapplies to similar services in other sectors (e.g. regulations onfinancial contracts), and specific electricity regulation is notgenerally needed. However, these activities have generally beenperformed or controlled by the vertically integrated electricity

21


1. For instance, consumers with flat loads may be offered lower prices than consumers whose peak demandcoincides with the system’s peak load. Retailers can make a profit from adjusting prices for individual consumers.However, as time of use metering is becoming widespread, opportunities for profitably performing this activity tendto disappear.

monopolies. As a consequence, the unbundling and liberalisation ofthese services is often part of electricity market reform.

The disaggregation of functions and the introduction ofcompetition is reflected in the changing shape of the market andmarket participants. Under competition the vertically integratedutility gives way to a number of different and more specialisedmarket players including generation, transmission, distribution andsupply companies. The unbundling of the various functions, oftenmandated or facilitated by regulations, is a key factor in thedevelopment of these new markets (unbundling considered inmore detail in Chapter 5).

What are the Benefits of Reform?

Reform primarily reflects a concern that economic efficiency is notas good as it might be in the ESI, and hence prices to consumersare higher than what they might be.The inefficient performance ofthe old regulatory framework has given cause for concern.Widespread excess generating capacity, unexplained national andinternational cost differentials (e.g. between plants or betweencompanies), and persistent international (or inter-state) electricityprice differentials imply that there is scope for improvement.Inefficiencies have become more obvious and relevant in thecurrent context of slower demand growth and globalisation.

The main objective of ESI reform is to increase economic efficiency.This requires minimising the cost of electricity supply and ensuringthat electricity prices are in line with costs. Experience from otherindustries shows that competition is the most effective way toestablish sustained incentives to keep costs and prices down.Undercompetition, productivity grows, costs and prices decrease, andinnovation and product diversity flourish. The expectation is thatthese benefits will also result from the introduction of competitionin the ESI.The largest expected benefits of reform in the ESI are:

22


■ Lower prices resulting from competition: Competition puts adownward pressure on the profit margins of generators andsuppliers and provides an incentive to reduce costs.As a result,electricity prices under competition tend to be lower.

■ Lower prices resulting from increased electricity trade: Reformfacilitates inter-system competition and trade in electricity,resulting in a better allocation of resources and, ultimately, areduction in the cost of supplying electricity. This is animportant benefit of reform in many regions. In the EU, theElectricity Directive aims to develop the EU internal electricitymarket by integrating the national EU electricity systems. In theUS and Canada, reform is expected to reduce the large (andinefficient) electricity price differentials that exist acrossregions. For instance, the average price in 1995 ranged fromabout 4 cents per kWh in Oregon to more than 10 cents perkWh in California and New York. In Australia, the NationalElectricity Market facilitates inter-state trade and therefore hasa similar effect on price differentials. Even though no preciseestimates are available, some preliminary research suggests thatthe gains from inter-system electricity trade may be substantial(IEA, 1995).

■ Savings in investment costs: better investment decisions areexpected, particularly in generation, as investors assume therisks of their investments, and incentives to over-investcorrespondingly disappear. The large generation capacityreserve margins now observed in some countries, as large as50%, can be expected to adjust to more normal levels as thecosts of non-economical investments cease to be borne byconsumers. Reserve margins in the order of 20%, which areobserved in countries like the US and the UK, may becomemore common.

■ Higher labour productivity: better use of manpower isexpected, particularly in distribution activities, as increasedregulatory oversight, incentive regulation and, in some cases,

23


privatisation, build up pressure for a more efficient use ofresources. Experience from the UK and Australia indicates thatlabour productivity increases in distribution can be large,reaching levels as high as 40 to 50% in some cases.

■ Development of new energy services.

Competition also means that regulators do not need to provideregulatory incentives to promote efficiency. The evidence frommany regulated sectors suggests that regulation is an imperfectsubstitute for competition in providing incentives for efficiency.A number of factors seem to cause this, including informationasymmetries between the regulator and the regulated firms andregulatory capture.

Reform needs to be carefully targeted at the largest inefficienciesin electricity supply. Historically, power generation became the firsttarget of reform because it offered the largest potential forimprovement. Costly planning errors, leading to excess generatingcapacity suggested that efficiency could be improved. Generationaccounts for about half of the total cost of electricity and it doesnot exhibit significant economies of scale so that competition is apossibility.

More recently, distribution and retail supply have also becomeareas for reform. Experience with the corporatisation andprivatisation of distribution systems indicates that potential costsavings in distribution, which accounts for about 30-40% of totalcosts, could also be significant. Experience gathered in competitiveelectricity markets also suggests that consumer choice is needed inupstream activities and that a level playing field for competitionrequires eliminating distortions in end user tariffs. Many recentreform programmes aim to unbundle distribution networks fromsupply, introduce consumer choice and re-regulate distributionnetworks.

The remaining link, transmission, is also now an object of reform,even if for a different reason. The direct potential savings intransmission are relatively small — its share of total cost is just

24


about 10% — and there is limited scope for competition. Buttransmission can easily become a bottleneck for the developmentof competition in the other ESI functions. Electricity regulatorsaround the world are now struggling with the regulation oftransmission, including transmission pricing, governance andunbundling from other activities.

What are the Costs of Reform?

A monopolistic electricity network remains essential to supply alarge majority of end-users. In consequence, extensive resourceshave to be devoted to the regulation and restructuring of thenetwork. Regulating the grid is a costly activity. Resources areneeded both to develop new regulations and to implement them inareas such as pricing, allocation of access rights and antitrustenforcement. Regulation also commits the resources of marketplayers as they try to influence the design of regulations and enterinto legal battles regarding their implementation. In addition,a detailed regulation of the electricity grid has costs in terms of foregone economic efficiency: regulatory loopholes andimperfections seem almost unavoidable given that regulators havelimited access to information in key areas such as costs ortechnical constraints (e.g. capacity and availability of lines).

The unbundling of activities along the vertical supply chain alsoresults in transaction costs. A web of contracts and intermediariesreplaces the previously vertically integrated structures. Some ofthe costs of contracting and intermediation already existed or havetheir counterparts in the integrated model. However, the total costof these activities is expected to increase.

It has been suggested that structural policies, such as unbundlingand divestitures, may lead to foregone economies of scale and co-ordination but the evidence suggests that foregone economies maynot be significant. For instance, there are economies of verticalintegration that may be lost when transmission is unbundled fromgeneration. Generation and transmission can be substitutes for

25


each other since the services provided by a transmission line canalso be provided by a power plant located where the line woulddeliver energy. Unbundling generation from transmission has thepotential cost of yielding sub-optimal investment decisions ineither activity. However, these indirect costs of vertical de-integration seem small in the light of the international experiencediscussed in Chapter 3.

Likewise, the horizontal restructuring of the ESI does not seem tohave a significant impact on the internal efficiency of electricitycompanies. It is sometimes claimed (e.g. in antitrust cases) thathorizontal restructuring may have efficiency costs because offoregone economies of scale.While there is general agreement thata minimum efficient scale exists for each ESI activity, this minimumgenerally allows for several companies to compete within the samemarket. As an example, studies of generation costs for US verticallyintegrated utilities suggest that scale economies are exhausted forproduction levels in the 12 000 to 30 000 GWh range. Similarly,distribution costs (including end-user supply costs) in the US donot seem to decrease with size, at least for utilities serving 750 000customers and above. Other studies find that the hypothesis ofconstant returns to scale cannot be rejected and that economiesof scale at the generating unit level are exhausted at a unit size ofabout 500 MW2.

What is Driving Reform?

While the main expected benefit of reform is increasingefficiency, this is not a new policy objective and, therefore, it maybe suspected that other factors are triggering reform. Someexperts believe that country specific factors have a significantweight in promoting electricity reform. In the US, the significantgap between electricity prices and (long-term marginal) costs has been a key contributing factor in the increasing pressure

26


2. See Wolak (1997), Frankena and Owen (1994), Lee (1995) and Joskow (1987).

for reform, especially in high price states like California andMassachusetts3. In the EU, political pressure for the developmentof the internal European market is a major factor in the processof reform. In Japan, high electricity prices and low generatingcapacity utilisation, resulting from a very uneven load curve, areencouraging reform. And in a number of countries (e.g. the UK,Chile, Argentina) reform has been linked to privatisationprogrammes and a wider reform of the economy.

In addition, there are a number of shared factors that may beencouraging reform but whose actual weight is somewhatdisputed. Current technological trends reinforce the advantages ofintroducing competition into the ESI. Economies of scale inelectricity generation are not significant enough under the currentconditions, thus paving the way for competition among electricitygenerators. In addition, cheap and abundant gas supplies haveencouraged the development of gas fired plants that are efficienton a relatively small scale.

Increasing economic globalisation also encourages electricityreforms. In a closed economy, inefficiencies in any part of theeconomy can be more easily absorbed by other economic sectorsto the extent to which they are also shielded from competition,and excess costs can be passed on to consumers in the form ofhigher prices. In an open economy, industries exposed tocompetition cannot remain competitive if they pay more for theirinputs than their (“foreign”) competitors. Globalisation thuscreates extra pressure to increase efficiency in electricity supply.Also, a global economy helps reform by fostering the emergence ofinternational energy companies that have the resources, thewillingness and the dynamism to compete in the newly liberalisedelectricity markets, and to introduce new ideas.

Finally, the fact that reform has been successful in some countriesprovides an impetus for reform in other countries. Regulatoryknow-how can be imported, at least partly, and the biggestuncertainties and concerns that surrounded the first reformexperiments are no longer present.

27


3. Joskow (1997).

A REVIEW OF ESI REFORMSIN OECD COUNTRIES

A Brief History

The origins of the current wave of reforms in the ESI can be tracedback to the late 1970s.The first step came with a partial openingof electricity generation to new entrants. In 1978, the US adoptedthe Public Utility Regulatory Policies Act (PURPA), requiring theutilities to buy electricity from “qualified facilities”, mostly co-generators and small power producers. Four years after thePURPA, in 1982, Chile enacted a law introducing some competitioninto electricity markets by allowing large end users to choose theirsupplier and freely negotiate prices. A second step came with theestablishment of explicit market mechanisms to determine thedispatch of generators and the wholesale price of electricity,thereby permitting competition between generators. The Englandand Wales electricity market, or “pool”, established in 1990, wasthe first such market mechanism. It was followed in 1991 byNorway which established a competitive electricity pool.This poolwas extended in 1996 with the incorporation of Sweden, andNordPool was formed, which now also includes Finland andDenmark.The National Electricity Market of Australia was createdin 1997 from the merger of the Victoria Pool, in operation since1994, and the New South Wales Pool, established as a daily pool in1996. In New Zealand, after ten years of reforms, a voluntaryWholesale Electricity Market was established in 1996, following thecorporatisation of generation in 1987, and the corporatisation ofdistribution and the introduction of consumer choice in 19944.

Competitive power exchanges started operation in 1998 in Spainand, within the US, in California and the “Pennsylvania-New Jersey-Maryland Interconnection”, opening the way for a number of

29

a review of esi reforms in oecd countries3

4. Other more tightly regulated pools were also created during the 90 (e.g. Argentina, 1991; Colombia, 1995;Alberta (Canada), 1996).

regional electricity markets within the US such as the New YorkPool and the New England Pool. In the Netherlands, theAmsterdam Power Exchange began operation in 1999. In the UK,an in-depth reform of the England and Wales Pool, known as NETA(New Electricity Market Arrangements) was approved in 1999 andimplementation is expected in early 2001.

In parallel to the development of wholesale markets, electricitymarkets have been progressively opened up to end users. In somecountries and states all end users are legally permitted to choosetheir supplier (Norway since 1991; New Zealand since 1994;Sweden since 1996; Finland since 1997; California since 1998;England and Wales since 1999; New South Wales since 1999).Thereis also some degree of market opening in many other OECDcountries even if no organised electricity market has beenestablished.

Early Reforms

This section summarises the approach adopted by early OECDreformers, those for which an open market has been in operationfour or more years. They are the UK, Norway, Sweden, Australiaand New Zealand (more recent reforms are considered in thefollowing section).

Key elements of early reforms are summarised in Table 2 below. Allearly reformers took some action to separate potentiallycompetitive from non-competitive activities and to stimulatecompetition in the competitive activities. In essence, this meant theseparation or “unbundling” of transportation from other activities,and the introduction of third party access to the grid. Earlyreformers also introduced an organised wholesale market.However, many detailed aspects of reform varied significantly,including the separation — or not — of generation fromdistribution, the governance of the grid, transmission pricing, theorganisation of the wholesale market, ownership, and theinstitutional framework.

30

a review of esi reforms in oecd countries === 3

31


Industry Structure

Country Ownership Vertical Integration Common of Utilities of Generation Elements

/Distribution

UK Private No* Full Consumer Choice (Target)

Norway Largely Public Yes Unbundling of Generation from Transmission

Sweden Mixed Yes Regulated TPA to the Grid**

Australia Mixed No

New Zealand Largely Public No

Table 2

Key Characteristics of Some Early Reformers

* Distribution companies are allowed to own some generation assets.** Except in New Zealand where TPA is negotiated.

Wholesale Markets

Country Pool Participation Pool Capacity* Pool Pricing**Payments

UK Mandatory Yes Ex ante

NordPool Non-Mandatory No Ex ante

Australia Mandatory No Ex post

New Zealand Non-Mandatory No Ex post

* Capacity payments: payments made to generators in order to guarantee their availability to generateelectricity in case of need.

** Ex ante: pool purchasing price determined from scheduled demand and supply.Ex post: pool purchasing price determined from actual demand and supply.

■ Structural Policies

The purpose of horizontal restructuring is to mitigate marketpower in the potentially competitive functions in electricity supply,especially generation and end user supply. Many reformingcountries have adopted measures to reduce horizontalconcentration in generation and in end user supply. Thesemeasures include mandated divestiture of generation assets andthe split of generation companies at the time of privatisation,generally intended to reduce market power.There have also beensome mandated splits of regulated distribution companies. Thispolicy aims to establish a critical mass of competitors and oftenalso a sufficient number of “comparators” for regulatory purposes(e.g. for setting price caps in the UK). An added benefit ofdivestiture is to set the market value of assets, which helps tohandle the transitional issue of placing a value on stranded assets.

The feasibility of structural measures to increase competitiondepends on whether the utilities are publicly or privately owned. Ifthe starting point is a government-owned industry, a government-mandated split of companies may be feasible.This was the case inthe UK at the time of privatisation, where generation was split intothree companies and distribution was split into thirteen regionalcompanies. In Australia, the Victorian monopoly has been privatised

32


Institutional Framework

Country Electricity Regulator Regulator Also Enforces Competition Law

UK Independent No

Norway Ministry No

Sweden Independent No

Australia Independent Yes

New Zealand No Specific Electricity YesRegulator

and disintegrated into five generators and five distributioncompanies; and, in New South Wales, restructuring has resulted intwo generation and six distribution companies, all kept underpublic ownership. In New Zealand, the dominant electricityproducer was separated in 1996 into two companies accountingfor approximately 60% and 30% of the generation market. In 1999,to increase rivalry among generators, the government divided thedominant company into three competing state-owned companies,and sold off the other state owned generator, reducing the shareof the two largest companies to about 53%.

Similar policies have been adopted in countries outside the OECD.For instance, in Argentina, privatisation transformed the threestate-owned generators, seventeen regional distributors and anumber of smaller regional generators into a verticallydisintegrated structure in which no generator has a market sharelarger than 10%.

Government-mandated divestiture policies may be difficult toimplement when there is private ownership of the utilities.However, restructuring policies may still be implemented on the basis of either antitrust laws or financial incentives. In 1997,the UK electricity regulator encouraged divestiture of 6000 MWof generating capacity owned by the two largest generators,threatening to refer them to the Mergers and MonopoliesCommission. In California, financial incentives, in the form of ahigher allowed rate of return on certain investments, have beenused to induce divestiture of 50% of the fossil-fuel generationassets owned by the largest (privately owned) utilities. However,the three utilities involved voluntarily decided to divest all butnuclear generation.

In some countries, horizontal restructuring has occurred implicitlythrough the integration of national (or regional) electricitymarkets. Integration of various national markets yields a lessconcentrated market because it increases the number of playersand the relative size of all players is smaller in the larger market. Inthe NordPool member countries, horizontal restructuring has

33


taken the form of unrestricted international trade.The AustralianNational Electricity Market integrates previously separated statemarkets.The EU aim to create an internal market in electricity goesin the same direction.

Horizontal restructuring has led to a significant decrease in marketconcentration. However, concentration may still be too high insome markets. Table 3 provides measures of concentration inpower generation.

Table 3

■ Ownership

Another element linked to structure is ownership. There is noclear ownership pattern among the early reformers (see Table 2).This mixed configuration of ownership can also be observed in theunreformed ESI. Nevertheless, there is a worldwide trend towardsmore private ownership due to both privatisation programmes andto the entry of new privately-owned competitors.The sequence inwhich liberalisation and privatisation measures have been adoptedalso varies among countries. Privatisation of electric utilities haspreceded liberalisation in the UK, while liberalisation has precededsome partial privatisation in the Nordic countries. In Australia,some privatisation has taken place both before and after

34


Market Market Share of the Two Largest Producers (C2)

1996 1998

UK (England and Wales) 55 41

NordPool 40 35

Australia (National Electricity Market) 40 36

New Zealand 90 53

Table 3

Horizontal Concentration

liberalisation. In New Zealand, while most of the industry hasremained under public ownership, some privatisation has occurredwith more expected in the near future.

■ Regulation

Early reforms sought to establish market mechanisms to determinethe dispatch of generators and the wholesale price of electricity.This can be done by either an organised wholesale market (alsoknown as a power exchange, or pool) or by buyers and sellers of electricity engaging in bilateral trade. A mandatory powerexchange jointly manages system operation and dispatch in Englandand Wales (the introduction of new non-mandatory arrangementsis in prospect) and in the Australian National Electricity Market. Anon-mandatory power exchange that establishes a merit order ofbids among participants5 and a separate system operator exist inthe Nordic countries and in New Zealand (see Table 2).

The operation of electricity pools varies. Some incorporatecapacity (or availability) payments to generators in addition toelectricity payments while others only pay for electricity. Poolpurchasing prices are sometimes determined by scheduled supplyand demand, in which case selling prices have to add a correctionfor the cost of balancing actual supply and demand. In other casesthey are determined by actual supply and demand and thenpurchasing and selling prices coincide (see Table 2).

Pricing of access to and use of the grid also had to be tackledalongside industry restructuring and the establishment of awholesale market. Without effective regulation of transportationthese other reforms would have been quite ineffective. A variety oftransmission pricing systems have been developed among the earlyreformers. In Europe — Norway, Sweden, England & Wales andFinland — pricing is based on a postage stamp tariff that is notsensitive to congestion.Within each of these systems, congestion

35


5. As will be seen later, in other countries there is no power exchange (e.g. Germany) so that all trade is“bilateral”.

leads to re-dispatching. In Norway and Sweden, transmission pricesvary according to a loss factor — intended to capture marginallosses — that changes from region to region. In New Zealand, anodal pricing system is in operation so that prices of electricity andtransmission are calculated to equate supply and demand at eachnode of the grid. In Australia, a zonal pricing system has beenadopted. Prices are set to equate supply and demand in each of anumber of pre-specified zones. Congestion between regions isaddressed by operating separate markets in each region. All thesepricing systems set a fixed charge to raise the required level ofrevenue.

Incentive regulation of grid activities has been introduced in somereforming countries. Traditionally, regulated activities have beensubject to cost of service regulation so that the revenue of aregulated company is set to cover costs, including a competitivereturn on investment. Under incentive regulation, allowed revenuesare still determined, initially, to cover cost of service. However,regulated companies are allowed to keep a fraction of any costreductions they achieve. In the UK, incentive regulation has beenimplemented through “CPI-X” (Consumer Price Index-X) pricecaps. Under this arrangement, the remuneration of grid activities isupdated yearly by a factor equal to consumer price inflation(measured by the CPI) minus a pre-specified X factor; currently theX factor is 3% for distribution and 4% for transmission. In addition,there have been one-off cuts, in the order of 20%, in transmissionand distribution charges. Overall, revised price controls haveresulted in substantial cuts in network charges. Transmissionrevenues in some Australian states are also subject to CPI-Xregulation.

Recent Reforms

Many other OECD countries are now embarking on reform.Reform is now underway in all the EU member countries, the US,Canada and Japan.

36


■ European Union

The Council of the EU adopted a Directive on the internal marketfor electricity (EC 96/92) on 19 December 1996 (EC, 1996). EUmember states, with some exceptions, implemented the Directiveinto their national laws by 19 February 1999. Ireland and Belgiumhave one additional year, and Greece has two additional years tocomply with the Directive.

Under the Directive, increasing shares of electricity markets mustbe opened to competition, based on size of user. For 1999, thegroup of largest users, accounting for at least 26.48% of the markethad a choice of supplier.This percentage increases to 30% in 2000and 35% in 2003. In practice, the minima mean that only large usershave the opportunity to choose their supplier, although memberstates can go — and often have gone — further.

A number of countries have opened, or will open, their marketsbeyond these limits.As of August 1999 there is full market openingin Finland, Sweden, the UK and Germany. Denmark will be in the

37


% of the national market open Minimum size ofto competition (consumption Eligible Consumers

Date* of eligible consumers relative (EU Average)to National Electricity GWh per year

Consumption)

February 1999 26 40

February 2000 30 20

February 2003 35 9

* Exceptions: Belgium and Ireland had one additional year to comply with the Directive. Greece has twoadditional years.

Table 4

Market Opening under the EU Electricity Directive

same situation by 2003, and Spain and the Netherlands by 2007or earlier.The remaining EU countries have not announced plansto go further than required by the Directive. Current andplanned market opening in EU member countries is summarisedin Figure 1.

Access to the grid is via a transmission system operator (TSO)who must be a separate business from the generation anddistribution businesses. Several EU countries (Austria, Belgium,Denmark, Ireland, the Netherlands and Portugal) have met thisobligation by requiring the TSO to be a legally separated companythat remains under the control and ownership of the incumbentutility.An even weaker form of separation (managerial separation)has been adopted in France, Germany and Greece. In somecountries, however, transmission is not vertically integrated with

38


Figure 1

Planned Electricity Market Opening in EU Countries

Source: IEA

the other electricity functions. There is operational separation inItaly and ownership separation in Finland, Spain, Sweden and theUK6.

EU member states can choose from three different procedures foraccess. Under regulated third party access, tariffs are regulated,published and available to all parties. Under negotiated third partyaccess, eligible consumers or generators/suppliers can negotiatenetwork access with the incumbent utility. Prices and access termsare agreed freely among them and are confidential. The systemoperators must be involved in the negotiations and must publish anindicative range of transmission and distribution prices on anannual basis.

The third possible approach is the single buyer system, where adesignated single buyer (expected to be the incumbent utility) sellsall electricity to final consumers. Eligible consumers are free toconclude supply contracts with generators/suppliers both insideand outside the incumbent utility’s territory. The electricitycontracted by an eligible customer is purchased by the single buyerat a price which is equal to the sale price offered by the singlebuyer to eligible customers minus a tariff for network services.

Most EU countries have chosen a regulated third party accessmodel as the primary form of grid regulation (see Table 5). TheEuropean experience with negotiated third party access (i.e. inGermany) has not been satisfactory. There are concerns thatnegotiated third party access is not effective in preventingdiscrimination and places a burden on the companies willing toaccess the network7.

The EU Directive provides two options for generating capacityadditions. Under the tendering procedure, the monopoly utilitydetermines when new capacity is required and conducts a tenderfor this requirement. Under the authorisation procedure, the

39


6. Definitions of the different forms of unbundling are provided in Chapter 5.7. EU Commission (2000).

timing of generating capacity investments is the responsibility ofindividual investors, provided that they meet criteria specified inadvance by the member state (e.g. environment, land use, publicsafety) for granting an authorisation to construct. Member statesmay also opt not to require a procedure and leave the addition tomarket forces.

The Directive also contains significant provisions that may delay oraffect the development of open markets. Member states mayimpose public service obligations to ensure “security, includingsecurity of supply, regularity, quality and price of supplies and...

40


Regulated Third Negotiated Third Single BuyerParty Access Party Access

Austria Germany* Germany (Transitional)

Belgium* Belgium*

Denmark Greece

Finland

France

Ireland

Italy* Italy*

Luxembourg

Netherlands

Portugal* Portugal*

Spain

Sweden

United Kingdom

* Combines the two systems.Source: EU DG XVII.

Table 5

Network Access in EU Countries

environmental protection”. Furthermore, “to avoid imbalance inthe opening of electricity markets” the Directive permits theimposition of reciprocity requirements.This means that a customerwho has choice in one member-state may be prohibited fromobtaining supply from a supplier in another member state wherecustomers of the same type do not have choice. In addition, theDirective also permits member states to impose a requirementthat up to 15% of fuels to be used in the generation of electricitycome from indigenous sources. A summary of how EU membercountries have complied with the main requirements of thedirective is provided in Table 6.

In some countries, the implementation of the Directive hasresulted in a major overhaul of the ESI. In Spain8, a competitiveelectricity market began to operate in January 1998. Main featuresof the new system are functional unbundling (i.e., separateaccounts) of distribution from both supply and generation; creationof an independent transmission system operator; arrangements fora non-mandatory power exchange set up to co-exist with bilateraltransactions outside the exchange; provisions to phase in fullconsumer choice over a period of ten years; and an extension ofexisting policies on coal and renewables. Overall, the long-termregulatory options are similar to those adopted by the Nordiccountries and some US states. However, there are sometransitional arrangements, such as the gradual opening of themarket, the introduction of a cap on wholesale prices and anagreement to gradually reduce end user tariffs over a five yearperiod, which are country specific.

In April 1998, Germany9 issued a new Act on the Supply ofElectricity and Gas. Utilities are required to provide negotiatedTPA, and all consumers are given free choice of supplier. Atransitional arrangement allows the Single Buyer Model to beapplied until 2005 in some municipalities. Until the end of 1999,

41


8. A detailed analysis is contained in the OECD/IEA Review of Regulatory Reform in Spain (1999).9. A detailed analysis is contained in the IEA Review of Energy Policies of Germany 1998.

there was an agreement between the utilities to set transmissionaccess prices on the basis of the distance between generator andconsumer. A uniform postage stamp tariff was charged for distancesless than 100 km, and a surcharge was added for longer distances.

42


Country Separation of Entry into Market Generation Transmission Generation Opening % Ownership

2000 2003

Austria Legal Authorisation 30 50 Mixed

Belgium Legal Autorisation 35 100 Privatised

Denmark Legal Autorisation 90 100 Municipal/coop

Finland Ownership Autorisation 100 100 Mixed

France Management Autorisation 30 35 Public

Germany Management Autorisation 100 100 Private/länder + municipal

Greece Management Authorisation 30 35 Public

Ireland Legal Authorisation 30 35 Public

Italy Operational Authorisation 30 40 Mixed

Netherlands Legal Authorisation 35 100 Municipal/ privatised

Portugal Legal Tendering 30 35 Mixed

Spain Ownership Authorisation 54 54 Privatised

Sweden Ownership Authorisation 100 100 Mixed

UK Ownership Authorisation 100 100 Privatised

Table 6

European Electricity Market Reform

Source: EU Commission and IEA.

Since the beginning of 2000, there is a postage stamp tariffcombined with a surcharge applied to all international and somedomestic transactions. These approaches to transmission pricinghave been widely criticised for being discriminatory and erectingbarriers to trade.They also raise concerns about the viability of anegotiated approach to transmission pricing. Despite theseproblems, the German reform has been successful in quicklypromoting price competition and significant price reductions.

In the UK a second wave of reforms is to be implemented in early2001. Under the New Electricity Trading Arrangements (NETA)there will be no central dispatch of generation. Instead, all tradingwill be made on a voluntary basis. In addition, capacity paymentswill be abolished, demand will play a more active role and thebidding structure in the new voluntary market will be reformed.The guiding principles of NETA are thus similar to those inspiringretail competition in other markets in Europe (e.g. NordPool) andNorth-America (e.g. California).

Despite recent reforms, cross-border transactions are a majorbottleneck in the development of the internal EU electricitymarket. The old pricing and capacity allocation mechanisms forinternational transmission lines are grossly inadequate in the newframework. For instance, cross-border tariffs discourage trade anddo not generally reflect the cost of transmission. Also, non-discriminatory TPA to the network is undermined by long-termcontracts and agreements granting access to cross-bordertransmission capacity to certain companies. The EU Commissionhas launched a process — known as the Florence process — toestablish common rules for cross-border transmission within theEU that are consistent with the development of the internalmarket. It is expected that a decision will be reached on this duringthe year 200110.

43


10. See Haubrich and Fritz (1999) for a detailed discussion.

■ US11

In the US, reform is taking place more or less simultaneously ingeneration and end user supply activities. On the generation side,emphasis is on ensuring open and non-discriminatory access to thegrid. On 24 April 1996, the Federal Energy Regulatory Commission(FERC) issued wholesale open access rules requiring transmissionowners to provide point-to-point and network services under thesame conditions they provide for themselves, and to separate theirtransmission and supply activities. In order to avoid discriminationin network access, FERC encouraged, but did not mandate, thecreation of Independent System Operators (ISO). These entitiesmanage and operate the transmission grid independently from thegenerators and other grid users without (necessarily) owning the grid. The operational unbundling of transmission provided by an ISO is an intermediate solution between full (ownership)separation of transmission and accounting separation. ISOs are notthe only solution to restructuring US wholesale electric markets.On December 1999, FERC issued Order 2000 that examines awide range of Regional Transmission Organisations (RTO),including Independent System Operators (ISO) and TransmissionCompanies or “Transcos”. A Transco generally owns transmission,rather than giving the ISO operational control of transmission.Order 2000 requires utilities to file a proposal for a RTO but fallsshort of requiring the utilities to establish a RTO. The US is stillstruggling with transmission pricing and how to ensure long-termplanning and expansion of the transmission system.

On the end user side, new regulation is concentrated on thefollowing: enabling supply choice for all consumers; leveling theplaying field for supply competition by means of unbundling andtransparency obligations imposed on utilities; and support of publicpolicy objectives and consumer protection.

44


11. A detailed account and assessment of electricity regulatory reform in the US can be found in the IEA Reviewof Energy Policies of the US (1998) and in the IEA/OECD (1998) study on Regulatory Reform in the US.

■ Canada

There are significant differences among Canadian provinces. OnlyAlberta and Ontario have firm plans to implement full retail accessin 2001. Both provinces have regulated third party access and

45


Electricity reforms in the United States are distinct from those in mostother OECD countries. First, they vary significantly from state to state.Thestate-to-state variation is greater than, for instance, in Australia, anotherfederal country, but is comparable to that among member states of theEU. Second, where end users get direct access to the electricity market,they typically all get access simultaneously (or over a very short period),unlike in Australia, New Zealand, and the EU member states, whereaccess is phased in over several years, and not always to all end users.Third, the reforms do not start from a unified, publicly owned system asthey do, for example, in France, New Zealand, and England and Wales.Having private rather than public initial ownership implies a muchgreater concern in the US about stranded costs. On the other hand, as inmany other countries, the reforms in the US have not includedprivatisation of publicly owned utilities.

A major part of the overall reform effort is to intensify competitionbetween generators to supply electricity.Among the requirements for suchcompetition in generation is non-discriminatory access to the transmissiongrid and ancillary services. Some states, such as California, are providingpowerful financial incentives to partially divest generation to owners fromoutside the present market. As an alternative to divestiture of allgeneration, California, the states of the Northeast and Pennsylvania, NewJersey, and Maryland have established “independent system operators”.Competition in generation also requires sufficiently unconcentrated

Box 1

Main Lines of Reform in the US

transmission is operated by separate entities.Alberta has set up amandatory spot market and a voluntary spot market is planned tostart operation in Ontario in 2001.

■ Japan

In 1998, the Government of Japan adopted a programme of partialliberalisation of its ESI. Large consumers, who use more than 2 MW

46


ownership of generating plants. In California, the divestiture of generatingcapacity was to multiple owners in order to deconcentrate generation.Independently run power exchanges have also been established in themore liberalised jurisdictions.

The second major reform element in the US is the promotion ofcompetition to supply all end users, which is allowed, but not required,under the Energy Policy Act of 1992. The Administration proposes thateach utility be required to permit all end users to choose their ownelectric power supplier by 1 January 2003, except where states or non-regulated utilities find, on the basis of a public proceeding, that analternative policy would better serve consumers.

The third major element of the US reform is the mitigation,measurement, and compensation for stranded costs. Mechanisms used torecover stranded costs include lump-sum exit fees and non-bypassablecharges on end users. Stranded costs are mostly attributed to investmentsin nuclear generation and in long-term power purchase agreementsunder the Public Utility Regulatory Policies Act of 1978. The range ofstranded cost estimates is US$70 billion to US$500 billion; an often-quoted likely mid-range is US$135 billion to US$200 billion.Source:“Regulatory Reform in the US” (OECD/IEA, 1998).

Box 1

Main Lines of Reform in the US (continued)

47


Figure 2

Status of State Electric Industry Restructuring as of May 2000

DC

1:Arizona, Arkansas, California, Connecticut, Delaware, District of Columbia, Illinois,Maine, Maryland, Massachusetts, Montana, Nevada, New Hampshire, New Jersey, NewMexico, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island,Texas,Virginia, and WestVirginia.

2: Michigan and New York.

3: Alaska and South Carolina.

4: Alabama, Colorado, Florida, Indiana, Iowa, Louisiana, Minnesota, Mississippi,Missouri, North Carolina, North Dakota, Utah,Vermont,Washington,Wisconsin, andWyoming.

5: Georgia, Hawaii, Idaho, Kansas, Kentucky, Nebraska, South Dakota, and Tennessee

Source: Energy Information Administration.

and take power at 20 000 volts or above, will be eligible to choosetheir supplier. These consumers account for about 30% of totalelectricity demand. In addition, there will be negotiated third partyaccess to the grid. Other related measures include a re-examination of the electricity rate (tariff) system, the introductionof a full scale bidding system for the development of thermalpower, and the removal and simplification of some administrativeprocedures and rules to ensure transparency in transactions.Reform was implemented in March 2000 and will be reviewedthree years later.

Performance

It is still too early to evaluate the performance — in terms ofcosts, prices, and global social benefits12 — of electricity reforms.The experience so far indicates that there are significantdifferences among reforming countries. For instance, in somecountries intense price competition has developed quickly while in others the initial impact of reform on prices has been modest.

Significant time series are only available for the UK. Since 1990, theproductivity of the industry has skyrocketed. Output rose by 8%from 1988 to 1995 and, in the same period, employment wasreduced by roughly 50%. Final electricity prices in the UK haveexperienced a substantial decrease over the 1990-1997 period. Inreal terms, domestic rates have decreased by 20%, equivalent to a9% reduction in nominal terms; non-domestic tariffs have fallen inthe range of 19 to 27% in real terms during the same period13.

Electricity generators have remained profitable, with a return oncapital employed that has stayed above 25% for the two largestgenerators in the 1993-99 period14.

48


12. Other key aspects of performance such as environmental impact and reliability will not be considered in thisbook.13. Littlechild (1998).14. Ofgem (2000).

The reduction of (end user) electricity prices in the UK hasresulted from both the adoption of regulatory measures (e.g. RPI-X caps on transmission and distribution charges) and competition.Decreasing pool prices15 have contributed to the decrease in finalprices.Time weighted pool selling prices have dropped nearly 10%in real terms in the 1990/91 to 1997/98 period16 reflectingsignificant reductions in the cost of power generation, estimated inthe 40-50% range17. However, the changes in costs have not

49


Figure 3

Market Share of Two Largest Generators in Selected OECDMarkets (1998)

Note: US and Canada not included as each comprises various separate markets. Share in New Zealand was reduced to53% in 99. Share in UK (England and Wales) was reduced to approximately 28% in 99.

15. The evolution of pool prices is described in the IEA Review of Energy Policies of the UK 1998.16. In nominal terms, pool prices rose during the initial four years of operation.This led to the imposition of a two year pricecap. Prices have since dropped.17. Ofgem (2000).

resulted in similar falls in pool prices which have remained wellabove the price that would make new investments economical18.This, and other related evidence, has been widely interpreted as asign of insufficient competition in the England and Wales pool. It hasbeen estimated that prices exceeded competitive levels by around20-25% in the England and Wales pool19.

A cost-benefit analysis of the UK experience provides estimates ofthe overall benefit of the 1990 electricity restructuring20.Dependingon the scenario considered it could be £6 or £11.9 billion (valuesdiscounted to 1995). This amount is equivalent to a permanentreduction of electricity prices of 3.2% or 7.5% from 1990 levels.The distribution of benefits among the different parties is unequalwith shareholders getting large benefits (£8.1 or £9.7 billion),government getting a smaller benefit (£0.4 or £1.2 billion) and(wholesale) power purchasers losing £4.4 or £1.3 billion,respectively21.

Evidence from other countries (see Table 7) may still be anecdotal:

■ In the Nordic countries, prices spiked in 1996 roughly at thesame time as NordPool began operation but this has beenattributed to unusually dry weather conditions during 1995 and1996. Since then, prices have receded to pre-95 levels. Forexample, in Sweden the average NordPool spot price in 1999was 119.42 SK/MWh, down from 120.49 in 1998, 143.77 in1997 and 260.01 in 1996.The current price level in NordPool isrelatively low compared to total electricity generation costs innew plants which are estimated to be in the order of 300-350SK/MWh for oil- gas- and coal- fired plants22. This steepdownward trend is not reflected in end user prices. Pre-taxdomestic prices rose around 3.5% in 1996, the year when themarket was open.At the end of 1999 they were roughly at the

50


18. Offer (1998), Ofgem (2000).19. Thomas (1999), Ofgem (2000),Wolfram (2000).20. Newbery and Pollitt (1997).21. The loss suffered by (wholesale) purchasers reflects that electricity prices have fallen less than costs.22. See Energy Policies of Sweden 2000 Review, IEA.

level of 1996 and were expected to decrease in 2000. Industrialprices have remained more or less stable in the 96-98 period.

■ In Australia, electricity prices decreased over 1% in 1998following the establishment of the National Electricity Market.A larger price drop was recorded in wholesale prices. In theVictorian pool, prices more than halved from $28.1 per MWhin 1995 to $12.5 per MWh in 1997. Prices remained low onaverage in the Australian National Electricity Market during1998 and 1999. This trend seems to reflect both the onset ofcompetition and the existence of large reserve margins ofgeneration capacity in the Australian market.

■ In New Zealand, final prices have not risen significantly after theimplementation of reform and some rebalancing has occurredbetween the domestic tariffs and other tariffs23. Average spotprices before and after the establishment of the New ZealandMarket have been fairly stable and have been below the price atwhich new generation capacity would become economical24.

■ In Spain, nominal retail prices have decreased by around 10% inthe 1996-1999 period following an agreement betweengovernment and industry to gradually decrease tariffs over afive-year period. Wholesale prices have remained stable overthe first two years of operation of the power exchange but havepeaked in the first months of 2000.Weak competition is oftenattributed to the high concentration of the Spanish ESI.

■ In Germany, an intense price war has been reported during1999 and early 2000, resulting in some significant pricereductions. According to Eurostat, prices fell in the 1996-1999period by 9.6% for industrial consumers and increased by 0.8%for domestic consumers. In parallel, a wave of mergers andacquisitions is reshaping the industry.This price trend seems to

51


23. Read (1998).24. Wolak (1997); New Zealand Ministry of development (2000).

reflect both the onset of competition and the atomisedstructure of the industry.

■ In California, regulated end user tariffs decreased by 10% at thetime of restructuring. Customer switching to new suppliers(presumably attracted by even lower tariffs) has been modest,particularly among domestic consumers. In some parts ofCalifornia, where the cap on tariffs was abolished in 2000, therehave been substantial price increases during the summer. It hasbeen estimated that energy purchase costs in Californiaaveraged about 14% above competitive levels during 1998 and199925. There has been significant price volatility in thewholesale market, with large price spikes during the highdemand summer months. Price increases have resulted from alarge demand increase, as well as from the investment slowdownthat took place during the years immediately before the openingof the market.The investment slowdown has been attributed tothe large regulatory uncertainty that existed at the time whenreforms were planned. Apparently, investment activity hasreturned to normal levels since the market opened in 199826.

Reliability and environmental impact are also critical dimensions ofperformance. Both impacts appear so far to have been modest and, if the right institutions are in place, there is no reason to fearthat this will change. However, regulatory uncertainty may have an adverse impact on reliability as illustrated by the case ofCalifornia27.While this is a transitional issue, it is a critical one.Theacceptability of reforms largely depends on the ability of thereformed ESI to sustain a reliable electricity supply.

An assessment of environmental and reliability performance has towait until much more information becomes available. Many factorsother than competition have a large effect on reliability and

52


25. Bushnell and Wolak (2000).26. CPUC(2000).27. Difficulties to attract investment in developing countries (e.g. Argentina, Brazil and India) have also beenattributed to the ambiguity and risk of the regulatory regimes in those countries.

environmental impact and therefore it is difficult to isolate theimpact of competition. For instance, the average CO2 intensity ofelectricity generation in the OECD countries has decreased since1990, and reforming countries participate in this trend. However,changes in relative fuel prices (e.g. lower gas prices) are probably a more important factor in explaining this trend than thedevelopment of electricity competition and trade. Likewise,reliability depends on reform policies and decisions taken beforethe onset of competition and unusual circumstances (e.g. severeweather or accidents). A longer time perspective will be needed toassess reliability under competition.

53


Table 7

Wholesale Average Prices (US$ per MWh)

Source : all data compiled by Offer (1998).

1995 1997

Australia (Victoria) 28.1 12.5

New Zealand (North Island) 19.1 26.0

Norway 15.6 17.9

United Kingdom (England & Wales) 39.7 40.5

WHAT MODEL FOR THE ESI?The review of national approaches in the previous chapter suggeststhe broad lines of an emerging new model for the ESI. This chapterdiscusses the emerging model and compares it with a number ofalternatives. The focus of this analysis is on how to maximiseefficiency in the ESI.

Retail Competition

The basic emerging alternative to the vertically integratedmonopoly is the retail competition model28. Most otherapproaches to reform can be described as a constrained version ofretail competition.The retail competition model has the followingcharacteristics:

■ Transactions between generators, end users and a number ofpossible intermediaries, including retailers, power exchangesand brokers, take place freely (within the constraints imposedby the network).Thus, on the demand side, end users are freeto choose their supplier and to negotiate their contracts; on thesupply side, generators can sell their electricity to any othermarket players.

■ Network activities and prices are regulated and, in particular,there are provisions to ensure non-discriminatory third partyaccess to the network, often including some form of separationof network activities from generation and end-user supply.

■ There is an independent system operator, which means that thesystem operator is not owned or, at least, not controlled by theowners of generation assets.

This model is the starting point for the organisation of theelectricity market in Finland, Norway, Spain, Sweden and Finland,

55

what model for the esi?4

28. Here we follow the most common terminology. See Hunt and Shuttleworth (1996) and Joskow (1996).The“retail competition” model is sometimes referred as the “bilateral contracts” model, particularly in Europe, but theprecise meaning of this term can vary.

some US states (e.g. California) and, with the implementation ofthe New Electricity Trading Arrangements, the UK. The basicstructure of the model is summarised in Figures 4 and 5. Retailcompetition combines deregulation, lifting constraints on thepotentially competitive activities in the ESI, with re-regulation ofthe network and related activities which remain monopolistic.

A comprehensive reform programme includes:

■ Structural reforms designed both to separate regulated andpotentially competitive activities and to promote competitionwithin the latter;

■ Institutional reforms intended to provide adequate frameworkconditions for the effective functioning of emerging competitivemarkets; performance regulation aimed at providing incentivesfor efficiency in the management of regulated activities;

56

what model for the esi? === 4

Figure 4

Monopoly vs. Retail Competition

57


Figure 5

Retail Competition: How Does it Work?

Figure 6

Other Approaches

■ Transitional arrangements to implement reform, including acalendar of reforms, and, in some cases, a plan to deal withstranded assets29; and

■ measures to ensure competitive neutrality of general policyinstruments (e.g. environment).

Alternative Approaches

Different industry configurations may be imagined combiningelements of the retail competition model with elements of themonopoly model. Only a few of these configurations have actuallybeen implemented.The first modern approaches to competition inelectricity markets were based on the portfolio manager model30,which is still widely applied in some regions of the world. In thismodel there is procurement competition, that is, the right to buildand operate generation assets is assigned competitively, usuallythrough an auction.Apart from this, all activities remain regulated,and monopoly utilities retain the obligation to supply consumerswithin their exclusive franchise areas with bundled retail electricityservice31. If the tendering process is open and competitive, thismodel provides incentives for cost efficiency in building andmanaging generation plants as generators can retain any costsavings. However, the portfolio manager inherits most of theweaknesses of the vertically regulated monopoly. There are noexternal (market) incentives to set end user prices efficiently.Investment risks are borne by end users instead of investors.There is no day-to-day competition among generators. Finally, theportfolio manager model is risky for end users because it locksthem into long-term procurement contracts that may eventuallyturn out to be too costly or otherwise inadequate.

58


29. The issue of stranded costs is discussed in the appendix at the end of this chapter.30. The term “portfolio manager” is somewhat vague. Most often it refers to generation procurement competition as definedin this paragraph. However, a monopolistic buyer of electricity not owning any generation plants is sometimes also called a“portfolio manager” (this manager could emerge in the context of the EU Directive single buyer option).31. See Joskow (1997).

Moving one step forward in the development of competition,generators may be allowed to compete against each other.Competition in generation is often labelled wholesale competition.It may require generators to sell through a power exchange; thisyields the mandatory pool model in operation in the UK andAustralia.Alternatively it may allow generators to sell directly to allor some end users or a number of intermediaries; this yields (fullor partial) retail competition. Often, the restrictions on consumerchoice are intended to be transitional arrangements, and areprogressively eliminated. Figure 6 summarises the portfoliomanager and the mandatory pool models.

Retail Competition Issues

Introducing consumer choice is costly, particularly for small endusers. This raises the issue of whether the benefits of retailcompetition, including the indirect benefits, outweigh the costs forall consumer groups.

The cost of metering may be a barrier to the development ofcompetition for small consumers. Currently, meters used forsmaller consumers do not generally have time-of-use meteringcapabilities.Time-of-use metering is needed, at least in principle, forthe unbundled billing of energy and grid services since energyprices (and possibly grid access prices as well) vary over time.Thecost of improving metering equipment is relatively high. Table 8shows some estimates of meter costs depending on the capabilitiesof the meter and the scale of its use. Rapid progress in informationtechnologies suggests that these figures will decrease in the futureand, even today, simple meters allowing for day/night differentiatedmetering can provide a relatively cheap proxy for the actual loadprofile. Metering also entails other costs such as those related tocompiling and processing information.

Compared with metering costs, the gross margin of electricitycommercialisation, i.e., the difference between the price paid by

59


end users and the price paid for energy and transportation, isrelatively small. For instance, the allowed gross margin ofcommercialisation in 1997 in the UK for below 100 kW consumerswas in the range of US$30-$38 per consumer per year. As aconsequence, even if commercialisation margins were to decreasesignificantly, smaller consumers may not generally find it profitableto pay the cost of improved meters in exchange for a reduction intheir annual electricity bill.

Reliable cost-benefit estimates of competition for small consumersare not yet available. Indirect impacts of competition on costreductions and tariff re-balancing are difficult to estimate. Thestart-up costs of metering and billing systems are also uncertain, asfigures provided by industry may be overstated. In the UK, theelectricity regulator estimated benefits of some £6 to £8 billionover a ten year period against costs of some £20 to £80 million peryear plus initial set up costs of between £150 to £520 million32.These figures suggest large welfare gains.When only direct impactsare considered and start-up costs are valued at an intermediate

60


Meter Type Functions Unit Cost Unit Cost 100 Consumers 50,000 Consumers

Meter Limited AMR* 175-300 75-300Modifications

Existing AMR* Load Profiling 250 100Electronic Meters

Advanced AMR* Load Profiling 600 500“Smart” Time-of-use ControlMeters

Table 8

Cost of Metering Equipment (US$)

AMR* = Automated Meter ReadingSource: J. King (1997).

32. Offer (1996).

value in the above range, one study33 has found that costs mayinitially exceed some £100 million per year. Once start-up costsare amortised, costs may approach benefits. These figures mayprovide lower bounds for the actual welfare gains.This study alsosuggests that there are large welfare transfers involved in theintroduction of competition, with consumers gaining significantlythrough lower prices, and with suppliers losing.

61


33. Green and McDaniel (1998).

Since April 1998, Californian electricity consumers may choose betweenDirect Access or their existing utility service. Direct Access is the purchaseof electricity from non-utility electric service providers. Furthermore,consumers below 20 kW may use load profiling to avoid new meteringrequirements.

There are concerns that retail competition will not easily extend to theresidential consumer segment. The low retail margins may notcompensate the costs of getting consumers to switch their electricitysupplier. Enron, one of the largest competitors in this segment, announcedin late April 1998 — three weeks after the Californian marketopened — its intention to abandon this market segment. On the otherhand, a number of electricity service providers remain active in thissegment.Apparently their strategy is to compete through the proliferationof value-added services (e.g. green energy) instead of competing in price discounts. As of August 1999, 1.5% of all electricity consumers,accounting for 11.6% of demand, have switched supplier in California,reaching an estimated 2.2% of all electricity consumers in February2000.

Source:Wiser, Golove and Pickle, 1998; Rivera-Brooks, 1999; and O’Rourke, 2000

Box 2

Competition for Small Consumers: the Case of California

An increasing number of countries are adopting “load profiling” toreduce the costs of extending retail competition to all consumers.The idea is that statistical inference procedures allow preciseestimates of the aggregate load of “many” small consumers. Even ifmetering and settlement procedures under load profiling arecomplex, requiring appropriate information technology, loadprofiling can be significantly cheaper than actual load metering.

Box 2

An Assessment

Retail competition aims to promote competition as much aspossible in the ESI, but acknowledges the need to continueregulating the network. Retail competition relies on end-userchoice together with competition among generators as the twokey forces disciplining the market. In older approaches, onlycompetition among generators was allowed to play a significantrole. By letting the demand side be an active market participant, theretail competition model seems to offer the best chances ofsuccess in addressing inefficiencies in the generation and end-usersupply segments of the ESI, whilst maintaining a strong regulatoryapproach to transmission.

Experience with approaches that do not fully deregulategeneration and end-user supply suggest that these are notsustainable in the long-term. Mandatory pools and partial marketopening require substantial regulatory involvement. Mandatorypools seem particularly vulnerable to manipulation or “gaming”,and partial market opening is likely to distort prices. As a result,captive consumers may subsidise eligible consumers. Theseproblems create pressure for removing restrictions to competitionin generation and end-user supply.

Full deregulation, in essence the removal of all electricity specificrules, is occasionally developed as an alternative means ofintroducing competition. Full deregulation lifts constraints on theprices of electricity and grid services, allows choice to all

62


consumers and does not impose any constraints on the verticalstructure of electricity companies. Under this approach, thevertically integrated structure of the industry remains unchanged.Network activities are not regulated. Market power and anti-competitive behaviour are controlled ex post (after the event),through the application of competition law. The negotiated thirdparty access models being implemented in Germany and, with amore significant degree of unbundling, in New Zealand, containmost of the elements of this approach.

Full deregulation is often the reform choice in other networkindustries, including telecommunications and air transport, and itcannot be ruled out in the ESI. However, it is complicated by thecentral role that the network plays in most electricity systems andthe consequent need which most reformers perceive to regulateit. In addition, many countries traditionally rely on ex anteregulation more than on ex post application of antitrust to pursuepolicy goals. Full deregulation in these countries would requirestrongly reinforcing their competition authorities and laws. Thisapproach could become a focal point of reform if changes intechnology or prices of generation reduce the importance of thenetwork in supplying electricity efficiently.

Partial or gradual reforms may have an important place in theoverall reform path. A gradual approach may favour the politicaland social acceptability of reforms. In addition, a partial reform maybe the only feasible option when governments seek to maintainsome direct control over the ESI, at least for some time. Reasonsinclude the desire to maintain or develop a technological optionsuch as nuclear, to keep a direct control over electricity prices orto promote national energy security beyond the levels that mightbe achieved by the market. Even when political constraints limit theextent and the potential benefits of reform, partial reforms canimprove upon the existing situation and can be built on in duecourse when political conditions allow. Examples of partial reformsare:

63


■ Institutional and procedural reforms to increase transparencyand to minimise undue influence in the regulatory process (e.g.establishing an independent regulator);

■ Vertical and horizontal separation of activities to increasetransparency and to facilitate regulation;

■ Dismantling subsidies and cross subsidies to eliminatedistortions in other economic sectors (e.g. end user tariffreform to eliminate cross subsidies);

■ Corporatisation possibly followed by privatisation of the ESI;

■ Incentive regulation of ESI activities and yardstick competition;

■ Limited competition in generation (e.g. procurementcompetition to create an external incentive for costminimisation on the utilities, and a mandatory spot marketbased on costs or, preferably, a generator’s bids to makedispatch decisions and to remunerate variable costs; and

■ Limited competition in retail supply (e.g. limited consumerchoice to facilitate some retail supply competition) or openingthe electricity market to international trade.

A number of case studies34 show that piecemeal reforms arepotentially costly. Regulatory mismatches expose governmentsand, ultimately, end users to substantial risks during thetransition. For instance, competitive generation procurementcoupled with monopolised supply may result in a build up ofpotentially stranded assets or potentially uneconomical long-term contracts. In addition, the eventual implementation ofdelayed reforms may be obstructed or blocked by early reforms.For instance, vertical and horizontal restructuring are much moredifficult after privatisation.Also, there is evidence that regulatoryintervention in the potentially competitive activities (e.g. througha mandatory pool) distorts the behaviour of market players.This

64


34. The UK experience has been extensively documented; see, for instance, Newbery (1995) and Green (1996).For an account of the early US experience, see Joskow (1997) and the references therein. See Rosenzweig andVoll (1998) for an analysis of reform in New Zealand, Argentina, Brazil, Chile and India.The case of Colombia isreviewed in Gray (1997). Bond (1998) points out some lessons from the Asian experience.

does not exclude gradualism nor flexibility in the implementationof reforms as long as reform progresses along a coherent plan.

Appendix

■ Transition Issues: Stranded Costs

This appendix reviews stranded costs, which are a significanttransition issue in some countries. An asset is stranded if its sunkcosts cannot be completely recovered under competition.Stranded costs are the difference between the sunk costs of anasset and the expected remuneration under competition (net ofvariable costs and salvage value). The potential for strandedgeneration assets seems to be significant in some countriesresulting from a number of factors, including excess generatingcapacity investment, lower than expected demand growth, andtechnological obsolescence. The concept of stranded costs hasbeen developed only recently in the context of the liberalisation ofthe US power industry. Explicit stranded cost payments have onlybeen set in the US and Spain.

Stranded costs raise three important issues. The first is thequestion of who should pay the stranded costs: consumers,taxpayers or investors? This is essentially a fairness issue. Sinceallocating stranded costs has distributive implications, no solutioncan be satisfactory to all parties. In countries evolving from apublicly-owned ESI, stranded costs can be amortised at the timeof flotation, and are implicitly absorbed by taxpayers.This does notdistort the market and leaves firms’ finances unaffected. However,this option may not be feasible when companies are alreadyprivately owned and taxpayers do not assume payment ofstranded costs. In this case, other arrangements betweeninvestors and consumers will need to be made. In addition, legalconsiderations may determine the allocation of stranded costs(e.g. regulatory commitments to finance certain investments andlong-term contracts). Stranded costs may also have implications

65


for the performance of the industry. In particular, stranded costpayments may have an unequal impact on competing electricitycompanies, distorting competition.

Second, there is the question of how to measure stranded costs.This is complicated by two interconnected features:

■ Competition creates incentives for firms to inflate theirstranded costs by either declaring more of their assets strandedor by claiming a low market value for their stranded assets. Ifpayment for stranded costs is made outside the competitivemarket, this strategy allows firms to improve their financialposition.The resulting inflationary pressure is sometimes calledthe snowballing effect of stranded costs.

■ The market value of a stranded asset under competition is notknown with certainty unless the asset is actually sold35. Theoption of selling the assets is sometimes a natural step in thereform process, namely when utilities are privatised at the onsetof competition or when divestiture is sought for verticalrestructuring to enhance competition. There can also be apolicy decision that each stranded asset be sold by means of an auction in which the former owners are not allowed toparticipate. Otherwise, the market value of the assets has to beapproximated following one of the several methodologies thathave been proposed.

If the stranded assets are sold, the measurement problems areresolved. However, this option may be unfeasible due to privateownership of the assets or to other reasons, in which casestranded costs have to be valued following ad hoc procedures.

66


35. The selling price will depend not only on the asset itself but also on the regulatory framework (e.g. whetherthere are capacity payments) and on the structure of the market (e.g. whether market power is expected to keepprices at high uncompetitive levels).This further emphasises the fuzziness of the stranded cost concept.36. For instance, a firm with many stranded assets may have a greater incentive to set low prices since this willresult in more stranded costs being recognised; a firm that owns no stranded assets will have “standard” pricingincentives.These asymmetries may negatively affect the performance of the market.

Third, related to the difficulty of obtaining reliable estimates ofstranded costs, there is the issue of when stranded costs should be valued. One possibility is to establish their amount at the onsetof competition. Another is to wait and link the assessment ofstranded costs to the actual value of the assets observed in themarket.The wait-and-see option imposes a financial cost on firmsdue to the increased uncertainty of future revenues. It may alsodistort prices and competition.The behaviour of the utilities will beaffected by the fact that market prices will be used by the regulatorto assess the magnitude of stranded costs36.Valuing stranded costsat the onset of competition does not create this distortion.

On the other hand, it has been argued37 that the wait-and-seeoption may result in less stranded costs being ultimately paid.Ex ante estimates of stranded costs may be biased upwards. Thisbias can be corrected once market prices and the actualperformance of the assets under competition can be assessed. Inaddition, stranded costs will most likely be reduced as thedepreciation of assets continues over time.Whether the wait-and-see option yields any real benefits depends largely on the politicaland institutional factors surrounding the bargaining process inwhich the amount of stranded costs is determined.These factorsvary from country and country. In any case, delays in the valuationof stranded costs can be costly and potentially distortionary and,therefore, have to be limited in time.

67


37. In the context of the Spanish and Dutsh reforms, for instance.

UNBUNDLINGThis and the following three chapters consider the four key areasof change that underpin an effective reform of the ESI aimed atmaximising competition.These are:

■ The separation of the functions that make up the ESI orunbundling (Chapter 5);

■ The organisation of wholesale electricity markets (Chapter 6);

■ The regulation of the transmission network and, particularly, thepricing of transmission (Chapter 7); and

■ The institutional and policy framework, which needs to beadapted to oversee the new market conditions effectively(Chapter 8).

Why Unbundle?

The main reason to unbundle is to avoid discrimination byvertically separating monopoly from competitive activities.Transmission, system operation and distribution remainmonopolies in a liberalised ESI. If these monopolies are verticallyintegrated with the competitive activities of generation and enduser supply, they have an incentive to use their monopoly poweragainst competitors.A grid monopolist can distort competition inmany ways. For instance, discriminatory access conditions, high ordiscriminatory access charges and “strategic” investment in gridaugmentation may put competitors at a disadvantage.

Competition law makes discrimination illegal in most countries,and discriminatory behaviour is punishable by competitionauthorities. However, this may not be a fully effective remedy tocounter discrimination. Showing that a certain practice isdiscriminatory may be difficult and costly, and requires the affectedparties to engage in lengthy antitrust procedures with noguarantee that they will win.

69

unbundling5

Vertical separation thus aims to limit the ability as well as theincentive of grid monopolies to distort competition. In particular,reforming countries have to deal with separation of:

■ Generation and transmission/system operation;

■ Generation and distribution; and

■ Distribution and end user supply.

A second and related reason to introduce unbundling is to improvethe effectiveness of regulation. Some degree of separation betweenregulated and competitive activities is needed to regulateeffectively. For instance, the regulation of transmission revenuesrequires, at least, separate and transparent accounting oftransmission. Stronger separation facilitates a more effective ring-fencing of regulated activities and, therefore, a more cost reflectivepricing of grid services.

Vertical Separation of Generation and Transmission

A transmission owner who also owns generation assets has theincentive to discriminate against the other generators and tofavour his own generating units. To the extent that competinggenerators have to access the network to deliver their electricity,the transmission owner also has the ability to discriminate bysetting high access prices, reserving transmission capacity for itsown generation units, providing unequal access to technicalinformation (e.g. changes in available capacity over time), orimposing abusive technical requirements. He can also enter intolong-term contracts that block transmission capacity or evenfavour a biased development of the transmission grid.

■ Forms of Separation

Vertical separation of generation from transmission and systemoperation activities may reduce or eliminate self-dealing or otherforms of discriminatory behaviour, but the architecture of

70

unbundling === 5

separation can be quite complex. Four basic approaches to verticalseparation have been proposed:

■ Accounting separation: keeping separate accounts forgeneration and transmission activities within the same verticallyintegrated entity. On this basis a vertically integrated entitycharges itself the same prices for transmission as it does othersand offers separate prices for generation and transmissionservices.

■ Functional separation: accounting separation, plus (1) relying onthe same information about its transmission system as theother market players when buying and selling power, and (2)separating employees involved in transmission from thoseinvolved in power sales.

■ Operational separation: operation of, and decisions about,investment in the transmission grid are the responsibility of an entity that is independent of the owner(s) of generation;however, ownership of the transmission grid remains with theowner(s) of generation.

■ Divestiture or ownership separation: generation andtransmission are separated into distinct legal entities withdifferent management, or operations, and there is no significantcommon ownership.

Some countries (e.g. Denmark) require corporate unbundling, i.e.,creating separate legal entities for generation and transmission,whilst preserving common ownership.This is a legal concept thatdoes not conform exactly to any of the four forms of separationdefined above. In practice, corporate unbundling may be similar toaccounting separation since it allows firms to share owners,management, staff and information while requiring a separateaccounting of the different functions.

The transmission function may be disaggregated into ownership ofgrid assets and system operation. Electricity trade through a powerexchange, which is sometimes bundled with transmission, can also

71

unbundling5

be performed separately. Transmission related activities may thusbe organised in many different ways depending on the form ofunbundling (from accounting separation to divestiture) and onwhich activities, among ownership, system operation and trade, areunbundled.Table 9 provides examples of the different approachesto unbundling transmission.

72

unbundling === 5

Type of Unbundling Examples Structure of Companies

Ownership Separate Finland, Norway, One TransmissionSeparation Power Spain, Sweden, Company (“Transco”)

Exchange UK (planned) owns and operatestransmission

One or several PowerExchanges (PX) facilitatetrade

No Separate England and Wales One TransmissionPower (until 2000) Company (“Transco”)Exchange owns and operates

transmission, and facilitatestrade

Operational Separate California One or several Grid Separation Power New Zealand Companies own the grid

Exchange System Operator (ISO)operates the grid

Power Exchange (PX)facilitates trade

No Separate PJM Grid CompaniesPower Interconnection own the gridExchange System Operator operates

the grid and facilitates trade

Functional and Accounting Austria, Belgium, Vertically Integrated Separation Denmark, Germany, Companies

Greece

Table 9

Organisation of Transmission-related Activities in Competitive Electricity Systems

■ Ownership Separation

Ownership separation solves nearly all concerns aboutdiscrimination because it eliminates the incentive as well as theability to discriminate. The weaker forms of separation limit theability to discriminate but do not eliminate the incentive to engagein discriminatory behaviour.

Imposing ownership separation of transmission on verticallyintegrated companies may be difficult due to legal obstacles oropposition from the utilities. Inducing divestiture by means offinancial incentives may be costly.This issue is particularly significantwhen the vertically integrated companies are privately owned, as inthe US and Japan.

Some analysts have expressed concern that ownership separationof transmission and generation may distort investment decisions ingeneration.The reason is that generation and transmission can besubstitutes for each other38. For example, an alternative to buildingnew generation capacity is to build a transmission line to bringelectricity generated somewhere else by an existing plant.Investment decisions that are limited to either generation ortransmission assets may not be globally optimal. For instance, weaktransmission links in a given area may discourage investments ingeneration that would be cost efficient in that area because of theproximity of a gas pipeline that could provide fuel cheaply.The costsavings for generation that would result from reinforcing the gridin that area may be ignored by a transmission company that is onlyconcerned with transmission investment.

However, the practical relevance of this issue may well be small.Ownership separation of transmission assets seems to haveworked satisfactorily for several years in a number of countries,including the UK, Sweden, Norway and Finland and no significantproblems have been reported. Regulatory measures, such asallocating the cost of congestion to the transmission company, can

73

unbundling5

38. They are also complementary as they are combined to supply electricity.

provide strong incentives to invest and eliminate congestion. Thisapproach, adopted for instance in Finland, may work even iftransmission investments are reserved to a monopolistictransmission company. Nodal prices, that reflect the true cost ofusing the network, also provide incentives for third parties toinvest and eliminate distortions. For price signals to be effective itis necessary that transmission investment be open to third parties(with appropriate limits on the participation of generators).

■ Operational Separation

Operational separation works through Independent SystemOperators (ISO) that operate but do not own the transmissiongrid. In effect, ISOs interpose themselves between transmissionowners and generators. ISOs may be effective in reducingdiscrimination, provided there are many competing owners ofgenerating units and transmission assets.

Operational separation raises other issues. First, incentives forefficient management of ISOs may be weak.The ability to controlISO management is reduced by the limited rights of the owners oftransmission assets. In addition, most ISOs are not-for-profit.Theirperformance will depend critically on the ISO governancestructure and how it combines objective independence andoperational expertise. Stakeholders can provide operationalexpertise but do not have the right incentives to act independently,as they tend to favour their narrow interests. For instance, the USexperience shows that weighted voting allowed abuse (in one ofthe old system operators, the NEPOOL), while requiringunanimous decisions made it difficult to reach decisions to correctmarket imperfections (in another of the old system operators, thePJM). Independent boards, on the other hand, lack operationalexpertise.

Existing ISOs have taken different approaches to governance. InCalifornia, the ISO is governed by stakeholders, but its functionsare restricted to the technical operation of the system. InNEPOOL and PJM, there is a two tier approach to governance, half

74

unbundling === 5

way between “all stakeholder” and “no stakeholder” governance. Inthis approach, an independent board, which has ultimate decision-making authority, co-exists with a committee of stakeholders,which makes decisions but the board does not review everydecision of the committee. The design of effective governancestructures for ISOs is still an open issue39.

A further issue is that incentives to invest in transmission may bedistorted to some extent.An ISO that controls, but does not own,transmission assets weakens the property rights of thetransmission owners.This may distort investment incentives as thevalue of the asset to investors may be reduced by their lack ofcontrol over it (owners can sell the assets but cannot control theiroperation). Experience with ISOs is still limited, so the significanceof property rights remains uncertain.

■ Functional and Accounting Separation

The two more limited forms of separation are relatively easy toimplement. Legal difficulties are minimal, opposition from industryis typically weak, and set-up costs are modest compared with thoseof setting up an ISO. However, the constraints on the ability oftransmission owners to discriminate are less effective even if theymay contribute to increase transparency. This implies thatfunctional and accounting separation, whenever adopted, needs tobe complemented by strong regulatory oversight, vigorousantitrust enforcement, and preferably both.These complementarymeasures are costly to administer and drain significant resourcesfrom the regulated parties.These costs have to be weighed againstthe benefits of a relatively simple and lower cost implementation.In addition, it is unclear whether increased vigilance by regulatoryand competition authorities would be enough to preventdiscrimination.

The foregoing analysis shows that there is no perfect solution:

75

unbundling5

39. See CPUC (2000).

■ Ownership separation is the only option that eliminatesconcerns about anti-competitive discrimination by transmissionowners.Thus, there is a strong presumption that divestiture oftransmission should be required in a competitive electricitymarket. However, if electricity companies are privately owned,ownership separation may be difficult to implement and, inpractice, other options may be more attractive.

■ Operational separation may be effective in preventingdiscrimination if there are many transmission owners. However,in order to promote efficiency, operational separation requiresthe development of sophisticated and still largely untestedgovernance structures.

■ Weaker separation forms (functional and accounting) require alarge and costly involvement of regulatory and competitionauthorities and may fail to prevent discrimination.

Vertical Separation of Generation and Distribution

Discrimination is also possible at the distribution stage.The ownerof distribution assets may favour his own generation anddiscriminate against other competing generators (e.g. charging highaccess prices to the distribution grid). The issues raised bydiscrimination in the provision of distribution services and thepossible approaches to deal with them are similar to thoseconsidered in the separation of transmission and generation40.However, operational separation of distribution is not consideredan option in practice because the costs of setting-up and operatingISOs are large relative to the size of a typical distribution area.

Divestiture is the only effective way to eliminate concerns withdiscrimination in distribution as it removes incentives todiscriminate.This is the approach taken in the UK and in some US

76

unbundling === 5

40. A difference is that the substitutability between generation and distribution is more limited.

states. Integration of generation and distribution is nonethelesspermitted in some reforming countries that have divestedtransmission (e.g. Norway, Sweden and Spain), but this may be anobstacle to competition.

Vertical Separation of Distribution and End-user Supply

Another possible form of discrimination arises when the owner ofthe distribution grid is also a competitor in the end user supplymarket. Abusive distribution pricing, cross subsidisation,unnecessary technical requirements (e.g. related to metering) andprocedural and implementation delays can be used to putcompetitors in the end-user supply market at a disadvantage.

The incumbent supplier (“utility affiliate”) benefits from asignificant competitive advantage vis-a-vis new independent(unaffiliated) entrants. Initially, the incumbent supplier covers theentire market. Thus, incumbents benefit from horizontal marketpower. Consumers may perceive risks and costs when switchingelectricity supplier. In addition, incumbents have an establishedreputation and recognition attached to the name and logo of theparent utility and they also have access to valuable informationabout their consumers. These structural barriers to thedevelopment of retail competition, particularly in the smallconsumers segment, reinforce the case for unbundling distributionand end-user supply.

The options for unbundling distribution and end user supply areessentially the same as those discussed above: separate accounts,functional separation (e.g. through the development of standards ofconduct to govern relationships between distribution companiesand their supply “affiliates”) and ownership separation. In addition,there are proposals for equalisation measures (e.g. in the US) tocompensate for the competitive advantage inherited by theincumbents.These include restrictions on the use of the name or

77

unbundling5

the logo of an utility by its affiliate and banning or restrictingaffiliate sales of electricity in the utility’s historic area41.

The actual approaches to unbundling distribution and end usersupply are typically functional or accounting separation, oftenimplemented on the basis of corporate unbundling. Ownershipseparation is required in New Zealand. The lack of unbundlingbetween the distribution and end-user supply businesses may slowdown the development of competition in end user supply, but astronger approach has not evolved for a number of reasons. First,implementing ownership separation is often difficult. Second, thedirect benefits of promoting competition in end user supply arerelatively small because it accounts for a small share of total costs(even if the indirect and long-term benefits are potentially crucialto reform success, as discussed in Chapter 4).Third, it is sometimesargued that there are vertical economies of integration betweendistribution and supply.

78

unbundling === 5

41. The issue of the appropriate relationship between a utility and its marketing affiliate has been addressed inthe US by several state regulators in the context of gas distribution. Standards of conduct in gas distributiontypically involve the prohibition of preferential treatment to the affiliate, structural separation of functions andrelated monitoring and enforcing procedures. In this vein, the California Public Utilities Commission requires non-discriminatory access to local distribution information; specifies separation standards (e.g. office space andinformation systems cannot be shared); mandates separate accounting; and establishes billing procedures in whichcost is separated by function and charges are the same for Direct Access and the utility service, except for energycosts (K. Costello, 1998).The much debated question of whether affiliates should be allowed to use the logo oftheir parent utility has been resolved by allowing them to do so.They are required, however, to inform their clientsabout their structural separation from the utility.

MARKETS

The development of competition in the ESI results in a largeincrease in transactions in electricity as well as in the developmentof related financial contracts.This chapter provides an introductionto organised electricity markets and related financial instruments.First, it provides an overview of organised electricity markets inOECD countries. Second, it discusses two key market design areas:bilateral transactions and capacity mechanisms.Third, the structureand function of financial contracts is analysed.

“Spot” Electricity Markets

The physical nature of electricity does not allow for a trueelectricity spot market, that is, a market for immediate electricitydelivery. Instead, transactions are scheduled some time in advanceof physical delivery (e.g. one day, one hour or five minutes inadvance). Imbalances between scheduled and actual supply anddemand that inevitably arise are handled following somepredetermined procedures, which may or may not be competitive.

Competitive pools or power exchanges are a substitute for a truespot market. In most existing pools, pool purchasing prices andscheduled supply are set by auction some time in advance ofphysical delivery. Pool selling prices are established by adding thecost of imbalances, ancillary services, and possibly other demandrelated charges such as capacity payments to the pool purchasingprice. Since prices are determined from scheduled supply anddemand, these are known as ex ante pools. Alternatively, there are ex post pools, like the Australian National Electricity Market, inwhich prices are determined ex post from actual generatorschedules and demand. In an ex post pool, the pool purchasing and selling prices coincide. Information technology is quicklyshortening the period in which system balancing needs to takeplace — though not eliminating such a need.

79

markets6

Box 3 below introduces the logical foundations of « spot »electricity markets.

80

markets === 6

In an electricity market, prices are used both to co-ordinate the decisionsof generators and electricity buyers, so that supply equals demand, andto ensure that these decisions are feasible given the physical constraintsof the system.

Prices

Spot prices for electricity would be set for each node of the grid. Anumber of cases will be considered in increasing order of complexity:

• Base case: In the simplest case,when there is enough generation andtransmission capacity to cover demand and transmission losses areignored, there would be a single price for electricity for each timeperiod:

Price of energy (P1) = Marginal cost of highest cost unit in operation

• At price P1 there is a generating capacity shortage: If settingprice equal to the marginal cost of the highest cost unit availablewould result in a generating capacity shortage, the above rule cannotbe applied.The price of energy has then to be increased in order todecrease demand. The price increase S needed to make demandequal to available generation capacity is the difference between themarginal cost of generation and the marginal benefit of consumption.The price of energy is then:

Price of Energy (P2)= P1 + S

S can be interpreted as the scarcity rent that covers the fixed costs ofgeneration.

Box 3

How an Ideal Spot Electricity Market Would Work

81

markets6

Therefore, S provides incentives for investment in generation.

• There are transmission losses: the price of energy (either P1 or P2, asit applies) would increase by a factor of (1 + Marginal Loss) at eachnode, reflecting that, in order to supply 1 kWh, it is necessary toproduce (1+Marginal Loss) kWh.Thus:

Price of Energy (P3) = P2 (1 + Marginal Losses)

• There are transmission constraints: the price at congestedconsumption nodes has to be increased so as to discourageconsumption; the price at congested injection points has to bedecreased so as to discourage consumption. The magnitude of theadjustment, known as the “shadow price” of the constraint is suchthat, at no point of the grid, supply exceeds transmission capacity.Theresulting prices are nodal electricity prices, discussed in detail inchapter 6.They include all the cases considered above as particularcases:

Nodal Price = P3 + Shadow price of the constraint at that node

Generation and Consumption

Because of the way wholesale prices are constructed, supply equalsdemand at each node and time period, and feasibility is ensured.Generators will produce energy if their marginal cost does not exceed theprice of energy at their injection point, but not otherwise. Consumers willbuy electricity up to the point where the price of electricity equals themarginal benefit of consumption.

Box 3

How an Ideal Spot Electricity Market Would Work (continued)

Power Exchanges: Review of theInternational Experience

This section provides some examples of the actual organisation ofcompetitive electricity markets. A summary of how trading isorganised is provided in Table 10. The review of the internationalexperience shows that pricing and scheduling mechanisms widelyvary. For instance, bidding can be iterative or one-shot and mayallow for demand side bidding or not. Bids can be firm or non-firmand can be simple, containing just a price per kWh, or may includeseveral terms. Prices can be determined ex ante or ex post and mayinclude capacity payments to generators.There can be pool priceceilings or other constraints on bidding behaviour; and transactionscan be settled in a number of ways. More generally, electricityexchanges differ in the degree to which optimisation by marketplayers — as opposed to optimisation by the “exchange” — isallowed.

Knowledge of the relative performance of different rules is stillquite limited but can be expected to improve in the near future as the number of power exchanges is increasing rapidly. Twoexceptions are the following:

It has been found42 that “electricity spot markets with mandatoryparticipation...tend to have more volatile prices than systemswith voluntary participation”. This study also considers marketstructure and technology. Price volatility is higher in fossil fuel-based systems than in hydroelectric based systems; and pricestend to be lower, but more volatile, when the companies areprivately owned than when they are publicly owned.

A comparison of the performance of two US markets43 —California Power Exchange and “PJM Interconnection” — finds thataverage prices and price differentials between constrained zones

82

markets === 6

42.Wolak (1997).43.Van Vactor (2000).

were similar in the two markets in the period until April 2000, butprice volatility was higher in PJM.

■ England and Wales Pool

In 1990, the ESI in England and Wales was unbundled into threegeneration companies, one transmission company and twelvedistribution companies. All these companies are now privately (andseparately) owned. Generation and, gradually, end user supply were

83

markets6

Market Participation Demand side Simple Pricing** Capacity Integrated Bidding bids* Mechanisms Dispatch***

England Mandatory No No Ex ante Yes Noand Wales

NordPool Voluntary Yes Yes Ex ante No No

Australian Mandatory Yes Yes Ex post No Partially NEM Integrated

New Voluntary Yes Yes Ex post No YesZealandElectricityMarket

Spanish Voluntary Yes No Ex ante Yes NoElectricity Market

California Voluntary Yes Yes Ex ante No NoPX (US)

PJM ISO Voluntary No No Ex post Yes Yes(US)

Table 10

Organisation of Electricity Exchanges in the OECD Area

* “Simple” means that bids are price-quantity pairs; “not simple” means that prices may have additio-nal terms.

** Ex ante means that prices are calculated for scheduled supply and demand; ex post means thatprices are calculated for actual supply and demand.

*** Integrated dispatch means that the system optimises joint use of generation and grid resources;otherwise there is unconstrained dispatch that ignores possible transmission constraints.

opened to competition, with the smallest consumers getting theright to choose their supplier by 1999.

The centralised wholesale market is the England and Wales pool. Itis mandatory: generators are obliged to sell their production to thepool and electricity buyers to buy from it.The pool sets prices forenergy for each half-hour period on the basis of a daily, one-dayahead auction. Generators submit bids specifying the capacityavailable for the next day and the price at which it is willing to selloutput from each capacity unit. Bids are fixed for the day so thatthe same price bids apply to all half-hour periods). With somelimited exceptions, there is no demand side bidding. Bid pricescontain several terms, such as a fixed start-up rate, a no-load ratefor each hour that the unit is running at its technical minimum, andvarious energy rates for different loads.The pool combines the bidsto construct an unconstrained merit order of generating plantsthat minimises the cost of serving the scheduled demand for eachperiod. Price bids are firm, but capacity bids can be withdrawn upto the moment of operation.

Scheduled generators receive the Pool Purchasing Price defined asthe system marginal price plus a capacity payment. The systemmarginal price is defined as the price of the highest bid needed tocover scheduled demand (where prices for start-up and no-loadare averaged and added to energy prices).

The capacity payment is intended to reward generators foravailability, thus providing “security of supply”. This value is fixedadministratively. Capacity payments are defined through a complexset of rules that ultimately aims to reflect the expected cost to theuser of a supply interruption or the value of capacity.This value iscalculated as the product of two quantities: the Value of Loss Load(VOLL) measured in pounds per kWh, and the Loss of LoadProbability (LOLP). VOLL is set administratively as there is nodemand side bidding from which the actual figure could be inferred.LOLP is set to take into account how much capacity is availablerelative to forecasted demand. It is higher when capacity is scarce.The amount (LOLP x VOLL), measured in pounds per kWh, is

84

markets === 6

charged on all energy sold and paid to all capacity that has beendeclared available but has not been scheduled.The size of capacitypayments varies significantly depending on available capacityrelative to demand as measured by LOLP values.

The unconstrained merit order may not be feasible due tonetwork capacity constraints ignored by the pool. If needed,the system operator calculates a constrained merit order.“Constrained on” units are paid their bid price plus the capacitypayment, and “constrained off” units receive the pool purchasingprice minus their bid.

Electricity buyers pay the pool selling price, defined as the poolpurchasing price plus the uplift.The uplift is the cost of the variousservices provided by the system operator, such as ancillaryservices, reserve and constraints costs, plus transmission losses.

A financial market runs in parallel to the pool. Contracts forDifferences may be used by generators and buyers of electricity tohedge the risk of price fluctuations.

There are plans to reform the wholesale trading arrangements inthe England and Wales pool.The reform proposals released by theElectricity Regulator in July 1998 and to be implemented in early2001, will abolish the mandatory pool, introduce a series ofvoluntary organised markets and allow for bilateral transactions.The bidding and price setting mechanism will be reformed on thebasis of three main building blocks:

■ First, the demand side will be incorporated into the pricesetting process. Incorporation of the demand side shouldencourage greater demand responsiveness and enable thesystem operator to balance the system at lower cost. It shouldprovide an incentive for suppliers to understand better theircustomers’ needs and enable them to manage demand better.

■ Second, offers and bids into the market will be firm. Participantswill be exposed to the costs and consequences of not meetingthe commitments in their offers and bids. Placing the risks and

85

markets6

responsibilities on those best able to deal with them shouldsharpen the incentives to manage and reduce risks.

■ Third, bids and offers will be in a simple form. Simple bids andoffers will be more conducive to transparency in price setting,incorporation of the demand side and management of dispatchcosts and risks.

These principles are to be delivered through the establishment ofa series of sequential markets that have much in common withother commodity markets. A forward market will operate long inadvance of real time delivery — perhaps a year or more ahead.Thismarket will enable parties to enter into contracts for the physicaldelivery of electricity. Closer to real time, a short-term bilateralmarket will operate.This will enable market participants to modifytheir long-term contractual position close to real time in order totake account of current information on matters such as theweather. From around four hours ahead, a balancing market willoperate. In this market, the system operator will buy offers ofincreased or decreased output, or decreased demand, in order tobalance the system. As with other commodity markets, there willalso be a derivatives market and a settlement process. There willbe no capacity payments in the new system.

The reasons for these proposed changes can be more easilyunderstood in a historical perspective. “The pool in England andWales was the first mechanism of its kind...[and]...it was developedin a process that gave considerable weight to the then existingarrangements...” (Offer, 1998). The proposals reflect a gradualconvergence of thinking in different countries about theappropriate organisation of generation in competitive electricitymarkets.

■ NordPool

NordPool is a voluntary electricity exchange open to traders fromNorway, Sweden, Finland and parts of Denmark. As of 1997, over40% of electricity trade in the area was handled by the pool.Thereis a spot market, Elspot, and a futures market, Eltermin, which deals

86

markets === 6

with futures contracts for up to three years ahead. Elspot is a oneday ahead auction market. Bids are made for each of the twentyfour hourly markets and consist of price-quantity pairs specifyinghow much the bidder is prepared to buy or sell at different prices.Supply and demand schedules are constructed from selling andpurchasing bids and, in turn, this determines a market clearingprice. Bids are firm, entailing a commitment to physical delivery orwithdrawal. All scheduled bids are settled at the market clearingprice.

A balancing or regulation market operates in each NordPoolmember country to manage transmission bottlenecks andimbalances resulting both from trade in the pool and from bilateraltrade. In Sweden, Svenska Kraftnät and in Finland, Fingrid make useof the countertrade principle. In other words, they balance themarket by re-dispatching generating units contracted for in aseparate balancing market. Svenska Kraftnät and Fingrid pay for thedownwards balancing of the surplus area and for the upwardsbalancing in the deficit area.The costs connected with the counterpurchase are regained through tariffs for transmission. In Norway,a split-the-market approach is used.The price of energy is reducedin the surplus area and increased in the deficit area until thetransmission demand is reduced to the capacity limit.The costs arerecouped from the market participants through the capacity fee inthe market settlement.

NordPool arrangements are the blueprint for other more recentlyorganised markets, including the Spanish and Californian marketsdiscussed below.

■ Australian National Electricity Market

The National Electricity Market (NEM) is a mandatory auctionmarket in which generators of 30 MW or larger compete.Wholesale market customers can also bid to the pool. Bids aresimple price-quantity pairs and ten such pairs can be submitted perday. Two additional “revenue bids” are also permitted specifying aminimum payment if the generator is forced to run below a certain

87

markets6

level. In principle, bids are firm. However, capacity bids can bealtered under certain specified conditions.

Bids are used to construct a merit order of generation and ademand schedule. Dispatch minimises the cost of meeting theactual electricity demand, taking into account « generic »transmission constraints. Thus, it approximates a fully integrateddispatch. Generation is scheduled according to this merit orderand regional spot prices are calculated ex post for each five-minuteperiod from actual supply and demand. Generators are paid thespot price.These arrangements eliminate the need for a separatebalancing market. Capacity payments are set equal to zero because,at the time prices are set, the probability of loss of load (LOLP) isalso zero.

A financial contracts market has developed in parallel to the NEM.Contracts for differences are bilaterally traded between the partiesto each arrangement. In addition, the Sydney Futures Exchange istrading two electricity futures contracts.

■ New Zealand Electricity Market

This is a voluntary market in which prices are determined on thebasis of simple (price-quantity) generator and consumer bids. Bidsare used to compute one day ahead optimal scheduling as well asreal time dispatch. Spot prices are determined from actual supplyand demand for virtually each node of the grid and for each half-hour period. Overall, the operation of the market is similar to thatof the Australian NEM, with the important difference that tradeoutside the pool is permitted in New Zealand.

■ Spanish Electricity Market

The general architecture of the Spanish market (Omel) is similar toNordPool, based on voluntary participation and firm bids.However, it incorporates an intra-day spot market that allowstraders a sequential adjustment of their trading portfolios at timesincreasingly closer to the time of operation. There are other

88

markets === 6

differences between the Spanish market and NordPool. Biddingprocedures are different since complex bids are allowed in Omel,and there are administratively set capacity payments.

The one-day ahead market sets prices for each of the twenty-fourhourly periods of the next day. Generators and buyers send bids tothe market operator who matches the bids. If the resulting basicdaily schedule is not feasible due to transmission constraints, themarket operator incorporates offers for congestion relief toestablish the definitive feasible daily schedule. Scheduled bids arefirm. On the day of operation, the intra-day spot market can openseveral sessions of trade (up to 24) for the remaining one-hourperiods of the day. Each session is similar to a one-day aheadmarket session.The outcome of this session is the basic intra-dayschedule, which subsequently becomes the final scheduling whenany possible modifications prompted by technical restrictions havebeen added into it.

Due to transitional issues, a financial contract market has not yetdeveloped.

■ US: California Electricity Market

The Californian Power Exchange (CalPX) conducts daily auctionsto allow trading of electricity in the forward day-ahead and hour-ahead markets. It is a voluntary pool and there is a large number ofcompeting power exchanges (known as Scheduling Coordinatorsin California). However, the major Californian utilities arecommitted to sell and buy only through the pool for the first fouryears of operation, until mid 2002.

The PX accepts demand and generation simple bids (price-quantity) from its participants, determines the market clearingprice at which energy is bought and sold, and submits balanceddemand and supply schedules for successful bidders to the systemoperator. It also submits bids for ancillary services, real timebalancing and congestion management. It is an energy only marketwith no capacity payments.

89

markets6

The trading procedures in the day-ahead market are as follows. Foreach hour of the 24-hour scheduling day, the PX constructsaggregate supply/demand curves. Their intersection determines the market-clearing price (MCP)44. The independent systemoperator (ISO) determines — based on all unit specific supply bidsand location-specific demand bids — whether there is congestion.If there is congestion, the ISO uses adjustment bids to submit an adjusted schedule to the PX. These adjusted schedules and ISO-determined usage charges become the foundation for zonal MCPs and for the final schedule submitted to the ISO. In theHour-Ahead market, bids are submitted to the PX at least 2 hoursbefore the hour of operation. The MCP is determined the sameway as in the Day-Ahead market.

■ US: PJM Interconnection Electricity Market

PJM (Pennsylvania/New Jersey/Maryland) is both a voluntary powerexchange and a system operator in the US. It operates a one-dayahead market in which generators submit offers that may include anumber of price terms. However, only one price bid per day can besubmitted, and dispatch is determined on the basis of these offers.PJM sets nodal prices for energy, also known as locational marginalprices. These prices are computed for the actual dispatch and,when transmission constraints are binding, prices are differentiatedby location.

PJM also operates a capacity market.This approach is followed byother US system operators such as Nepool and New York ISO, butnot in California.The capacity market results in capacity paymentsto generators, just as in the England and Wales pool, but thepayments are determined by the market instead of administratively.

All load-serving entities (LSE) are required to purchase installedcapacity (ICAP) in addition to energy. This requirement is a

90

markets === 6

44. Bids initially submitted into the day-ahead market auction need not be attributed to any particular unit orphysical scheduling plant. Such a bid is referred to as a portfolio bid. Portfolio bids that are accepted into the day-ahead market are then broken down into generation units.

function of the annual peak load served by each entity. ICAP isbilaterally traded between generators and LSE. LSEs that are shorton ICAP must pay a penalty to the system operator which is thenredistributed among generators. In addition, there is a monthly anda one-day ahead ICAP market created to facilitate ICAP trading.Generators with unsold ICAP are obliged to offer it in the one-dayahead market.

PJM also operates a Fixed Transmission Rights market to provideinsurance against price volatility caused by transmissioncongestion.

Trading Outside the Pool

There is a growing consensus that electricity trade should beallowed to take place outside organised markets. Bilateralcontracting is expected to be efficient since it is a standard, if notthe unique, way of trading in many markets. Bilateral trading is, bydefinition, more flexible than centralised pool trading since it mayco-exist, and it does in practice, with a non-mandatory pool.A non-mandatory pool also lessens concerns about discrimination and isa necessary condition for individualised pricing and provision ofsecurity and reliability, adapted to individual consumer needs.

In bilateral trade systems, market and system operation are oftenconducted by separate organisations. The system operator (SO)assumes most technical co-ordination functions for balancing ofthe system, including time of operation dispatch. The powerexchange (PX) organises trade among participants. This dualSO/PX structure reduces concerns that a joint SO and PX coulddiscriminate against traders engaging in bilateral transactions.Separate SO and PX exist, for instance, in NordPool, Spain andCalifornia, but not in PJM.

Bilateral contracting in electricity markets has, however, beencriticised on three counts:

91

markets6

■ First, bilateral contracting is not compatible with a centralisedoptimisation of dispatch. It does not guarantee dispatch basedon a merit order of bids or costs. However, as in most othermarkets, the lack of a central optimiser does not precludemarkets from being (potentially) efficient. This argument wasprobably a factor in the preference for a mandatory pool inearly reform models (England and Wales) but, since theNordPool experience has proved that a non centraliseddispatch may work efficiently, this argument is now lesscompelling.

■ Second, there are concerns that electricity prices to end usersmay not be transparent and/or pool prices may be distorted ifa large fraction of traders enters into bilateral contracts. Settingregulated tariffs to end users when the price of wholesaleelectricity is not clear may be difficult.The aim of reform is tolet market forces, instead of the regulator, set end-user prices.However, greater transparency may facilitate the transition to acompetitive market and a transitional requirement on largesuppliers to buy from the pool may be justified. Concerns aboutbilateral trading resulting in the exercise of market power havebeen addressed in California by means of transitionalarrangements that limit, but do not prohibit, bilateraltransactions.The utility distribution companies are required touse the PX during a four-year transitional period. For non-utilitybuyers and sellers of electricity, the use of the PX is optional.This strategy excludes the largest market players from enteringinto bilateral contracts without imposing constraints on smallerplayers.

■ Third, long-term bilateral contracts may facilitate the exerciseof market power, if market players already enjoy market power.In particular, bilateral contracts can result in an implicit form ofvertical integration between generators and distributors insystems where explicit vertical integration is not allowed. But ina sufficiently atomised market place with many potential buyers

92

markets === 6

and sellers, there would be little incentive to enter into bilateralcontracts at uncompetitive prices. This suggests that thepotential problem is market power in itself and thatuncompetitive bilateral contracts would only be a symptom.

Bilateral electricity trade has been allowed successively in theNordPool countries, New Zealand, Spain, the US and Germany.TheUK has proposed new trading arrangements that will allowbilateral trade. In California and (implicitly) in Spain, there are time-limited restrictions on bilateral trading imposed on the largestincumbent utilities.The trend is not universal, however, as Australiamaintains a mandatory pool. Power exchanges outside the OECDarea are often mandatory.

Capacity Mechanisms

Capacity payments are money transfers to power generators givenin exchange for making their generation capacity available.They arealso known as availability payments. With capacity payments, theprice of electricity paid to generators has at least two components.One component is related to actual energy production; the otheris determined by the generating capacity made available by thegenerators. Capacity payments may be set administratively, as inEngland and Wales and Spain, or through market mechanisms, as inPJM.

The broad policy objective of capacity payments is to inducegreater reliability of electricity supply than the market is expectedto provide. Capacity payments are ultimately intended to improvesecurity by encouraging a higher reserve margin, or by reducing thevariability of the reserve margin over time, or both. In a historicalperspective, however, capacity payments may be seen as theinheritance of planning methods applied before the introduction ofcompetition.

In principle, capacity payments could be expected to contribute togreater and more stable investment provided some conditions are

93

markets6

met. A key condition for capacity payments to induce investmentin the long run is credibility. Investors are not likely to modify theirinvestment decisions if capacity payments are perceived as atransitional measure that will be eventually abolished.The fact thatcapacity payments are being challenged and reviewed by regulatoryauthorities in both the UK and Spain suggests that the impact oninvestment decisions may be low. Another key condition is thatcapacity payments be reflective of the long-term value of capacity.If prices are, or can be, distorted they do not provide anappropriate signal for investment. In current practice, it is unclearwhether capacity payments reflect the economic value of capacity.

Regulated capacity payments, that is to say payments imposed bythe regulator, are introduced to modify the performance ofelectricity markets. Their introduction is linked to the belief thatthere may be some form of market failure resulting in a reliabilitylevel that is too low. Four main types of potential market failurehave been suggested45:

■ “Investment cycles”: It is sometimes argued that investment maygo through cycles. If, for instance, investors are “myopic” andtoo concerned with short-term price levels, then investmentcould go through pronounced cycles46. If, more plausibly,investors have a longer time horizon, then investment cyclesmay be less pronounced or negligible. Appropriate incentives,such as financial penalties in case of non-delivery, couldreinforce this. In practice, mitigating factors can compensate forany investment lags, such as new technology which is shorteningconstruction times, the scope for repowering existing plants,and the “demand-smoothing” potential of peak load pricing.

■ “Investing in reserve capacity is too risky”: Some electricitysystems, particularly those largely based on hydro generation,need substantial investments in reserve capacity that is only

94

markets === 6

45. Some observers (e.g. Ruff, 1999) argue that insufficient demand participation makes prices and revenuesinadequate. However, an artificially inelastic demand response should result in higher, not lower, prices.46. Ford (1999) and Ford (2000).

rarely and unpredictably used.These investments may be seenas too risky by investors. It is also argued that the prices thatwould have to be paid to make these investments profitable onthe “few” occasions the assets are actually used are too high.This may be a significant issue in some electricity systems.However, it is unclear that the appropriate solution wouldalways require regulatory intervention. For instance, generatingcompanies can diversify risks by owning a diversified assetportfolio; and both generators and suppliers can enter intofinancial contracts to reduce or eliminate risks. Increasedinternational trade in open electricity markets can also make asignificant contribution to diversifying the overall generationbase on which a country depends (for example, the NordPooldoes this in relation to Norway’s high dependence on hydro).

■ “High cost of capital discourages investment”: It has beenargued that the cost of capital for generating assets could beundesirably high in a competitive market due to the capital-intensive and long-lived nature of the assets.The result could beunder-investment and low security of supply.Another variationof this argument is that more capital-intensive technologiessuch as nuclear would be at a disadvantage in a competitivemarket. The cost of capital is certainly likely to be higher incompetitive markets. In the past, the return on investments inthe electricity supply industry was guaranteed by regulation, sothe owners’ investment risk was correspondingly low. Undercompetition, investment risk is shifted from electricityconsumers to owners, raising the cost of equity and possiblyinducing a decline in the share of debt used to finance theassets. Thus, the cost of capital for generation assets can beexpected to increase as it approaches its “normal” market levelreflecting the cost of capital in other, similar industries. Theadjustment in the cost of capital to reflect normal marketconditions should improve efficiency. In addition, variousfinancial techniques available in competitive markets allowcompanies to reduce their investment risks. Indeed, the

95

markets6

evidence suggests that investment in merchant plants47, whicharguably face a higher cost of capital, is not deterred by theopening up of markets to competition.

■ “Unsustainable prices”: It is sometimes argued that prices incompetitive electricity markets would tend to be below cost,discouraging investment.This argument rests on a misunderstandingof how competitive markets work and is not supported by theevidence. Competitive prices can be expected to cover allcosts. Wholesale electricity prices have not generally fallen tounsustainable levels. In some instances, there is concern thatprices are actually too high.

Beyond their potential impact on investment activity, capacitypayments may distort market performance in a number of otherways:

■ Capacity payments may induce inefficient strategic behaviour bygenerators. Many analysts have concluded that, in the Englandand Wales pool, there has been “gaming” by generators, that is,strategic manipulation of availability declarations to increase thecapacity payment.

■ Capacity payments generally increase wholesale and finalelectricity prices.

■ Capacity payments may distort competition and, particularly, theentry of new competitors, because they provide revenues toincumbent generators regardless of whether they are actuallyselling electricity.

■ Capacity payments treat all or most buyers and sellers ofelectricity homogeneously, regardless of their actual demandfor, or contribution to, security of supply. This is inefficient.Consumers range from small domestic users to large industrialcompanies and are likely to have widely different valuations forsecurity of supply that are not reflected in what they pay.

96

markets === 6

47. Merchant plants are those that must find a buyer for their output.

Capacity payments may fail to discriminate among investmentsthat make a significant contribution towards security andinvestments that do not make such a contribution (e.g. gas firedplants to cope with seasonal variations in a hydro-based systemversus additional hydro units).

■ Capacity payments can be used as a convenient way to mask the payment of stranded costs resulting from the pre-reformsystem.

There are regulatory alternatives to capacity payments including:

■ Obligations to ensure supply, with penalties on supplycompanies for non-delivery;

■ Direct ownership of generation units facing severe revenueuncertainty by the system operator (e.g. some peaking plantsthat are only rarely used, as in Sweden);

■ Monitoring of investment by the regulator with the possibilityof regulatory intervention if and when problems are anticipated;

■ A voluntary organised market for capacity to ensure bothtransparency and an efficient pricing of capacity; and

■ Interruptibility discounts to end-users.

Financial Markets

Financial contracts have a key role in providing insurance for marketplayers against price volatility in electricity markets. Prices incompetitive electricity markets quickly move in response tochanging demand and supply conditions and, as a result, are volatile.Price movements have a healthy effect in efficiently accommodatingsupply and demand but may have a negative effect on market players.Electricity contract markets reallocate price and quantity risks.

In the past, the risks involved in electricity supply were bundledwith electricity itself. One of the benefits of reform is that risk isunbundled from the provision of the basic good and can be, to acertain extent, separately and flexibly managed and priced.

97

markets6

Electricity contract markets are primarily financial markets and donot require specific electricity regulations. However, as in othercommodity futures markets, financial electricity contracts mayneed to be backed by financial penalties and guarantees to ensurethat the contracts are honoured.

It has been suggested that contracts may help to curb marketpower.The idea is that financial contracts reduce the incentives ofgenerators to set high prices in the wholesale market because theprice that a generator receives is set in the contract, not in thewholesale market48. Unfortunately, the evidence from the Englandand Wales pool suggests that financial contracts do not mitigatemarket power, at least not significantly.

How do electricity contracts work? Most contracts take the formof forward contracts, futures contracts, option contracts, or PowerPurchase Agreements (PPA). The first three are sometimesdescribed as financial contracts, because they do not need tospecify which plant will provide the power. PPAs are also calledphysical contracts, because they do specify the plant that willprovide the power. They are commonly used by independentpower producers (IPPs) selling to a monopolistic buyer but canalso be applied in other contexts.

Forward contracts are bilateral agreements to deliver energy at agiven price. They work in combination with a competitivewholesale market and are one of the simplest forms of derivativeinstruments that are used to transfer, or “hedge”, price risk. Theparties involved agree to a price today (the “strike” price) for“delivery” of a certain amount of energy later.The settlement of aforward contract can be made without physical delivery. If themarket price at “delivery” is higher than the strike price, the sellerof the contract compensates the buyer for the difference; if it islower, the buyer of the contract compensates the seller.A forwardcontract settled in this way is called a contract for differences.

98

markets === 6

48. However, since contract prices reflect expected wholesale prices, generators will generally make a higher profitif prices are “high”.

Futures contracts are analogous to forward contracts with theonly difference that they are standardised and traded in anorganised market.

Option contracts give the right, but not the obligation, to buy orsell electricity at a certain price.The price has two components: anoption fee equivalent to a kW charge payable when the contract issigned, and an exercise price to be paid for each kWh actuallydelivered. An option contract does not need the reference of aspot electricity price to be implemented. Compared with forwardcontracts, option contracts have the advantage (to the seller) ofpartially hedging quantity risk as fixed generating costs can becovered by the option fee.

Financial contracts are currently in use in electricity markets. Forinstance, forward and option contracts are extensively used amongtraders in the England and Wales pool, NordPool and AustralianNEM.These contracts are settled for differences. Electricity futuresare traded in NordPool, the US (NYMEX, PJM) and Australia(Sydney Futures Exchange). There is a plethora of far morecomplex and advanced derivatives which can be used to hedge allkinds of price risk.

PPAs are similar to the financial contracts described above but,since they specify the plant that will provide the power, theyrequire the generator to be excluded from any centralised pool. Incountries with a mandatory electricity pool these contracts areeither restricted or banned.

99

markets6

NETWORKS

Means and Ends of Network Regulation

Transmission and distribution lines provide the critical physical linkthat makes competition feasible.Thus open access to the networkand adequate pricing are essential for the development ofcompetition. In addition, to the extent that reforms also aim tointegrate previously separated (national and state) markets, theregulation of cross-border (or inter-state) transmission links is amajor issue in many countries.

This chapter considers the management of the grid. It focuses firston the role of prices in promoting efficiency. In the short-term themain regulatory issue is how to allocate scarce transmission anddistribution capacity efficiently. In the long-term, regulation isconfronted with providing adequate incentives for investment andcost efficiency without compromising the financial sustainability ofthe regulated companies. Further on, the institutional frameworkwith a focus on the role of system operators is considered. Anadditional pricing issue is how to allocate the large fixed costs of thenetwork to different users.This is discussed in the appendix.

Prices are the main tool in network regulation. The pricing ofnetwork services pursues a number of complementary objectives:

■ Allowing sunk investment costs to be recovered (financialsufficiency);

■ Providing adequate incentives for future long-term investment(long-term efficiency);

■ Providing adequate signals for efficient network operation, or in other words, for the allocation of capacity to managecongestion efficiently (short-term efficiency);

■ Avoiding discrimination among users of transmission(competitive neutrality); and

■ Promoting simplicity and transparency.

101

networks7

The third of these objectives (short-term efficiency) is particularlyimportant as well as difficult to manage. Efficient pricing for theallocation of capacity to deal with congestion may appear to be atechnical issue, with limited implications. However, it is a key issuein the overall design of efficient electricity markets because inaddition to its direct role in the allocation of transmission capacity,it affects the dispatch of generation units. The impact oftransmission prices on the distribution of revenues and profitsamong generators can thus be significant. In addition, transmissionpricing may have an impact on competition in the generationfunction, either facilitating or distorting it.

Many pricing methods have been developed in order to meet thevarious objectives.The average price level (or the revenue allowedto the regulated company) can be set to cover different cost andprofit definitions and adjusted to provide incentives for efficiency.Prices are usually charged through a multi-part tariff structure thatmay include fixed connection charges, as well as capacity andenergy charges that may depend on time of use. Each part of thetariff can be determined separately for each location, oralternatively for each user. In this way, many different combinationsof an average price level, charges by location and charges by user,are possible.

In practice, to achieve the different objectives transmission pricescombine a number of pricing and non-price methods. Table 11summarises the approaches adopted by a number of reformingcountries.

102

networks === 7

103

networks7

Country Average Pricing Dispatch Charges for Price Level by Location Recovery of

Sunk Costs

Australia CPI-X applied Zonal Integrated dispatch Fixed connectionto optimised (model differentiated plus postagedeprival value by regions) stamp energy

charges

Finland Cost Based Postage Stamp Unconstrained Market Energy BasedClearing + Countertrade Fees

New Zealand Optimised Nodal Fully Integrated Connection andDeprival Value Dispatch Capacity Charges+ Revenue cap

Norway CPI-X Postage Stamp Unconstrained Market Capacity Charge (Differentiated Clearing + “Split the for Peak Usageby Regions) Market”

Spain CPI-X Postage Stamp Unconstrained Market Capacity and Clearing + Energy ChargesCountertrade

Sweden Cost Based Postage Stamp Unconstrained Market Capacity(Differentiated Clearing + Chargeby Regions) Countertrade

US Historical Zonal Unconstrained Surcharge to End(California) Cost Market Clearing Users to Cover

Sunk Costs

US (PJM) Historical Nodal Fully Integrated Capacity Cost Dispatch Charge for Peak

Usage

UK (England CPI-X Postage Stamp Unconstrained Zonal Capacity & Wales) Dispatch + Charges

“Countertrade”

Table 11

International Comparison of Transmission Pricing

Source: Putnam, Hayes and Bartlett (1997) and IEA.

■ Overview

There are two main competing approaches to the pricing oftransmission services. There are non-transaction based, or pointtariffs, which are independent of the commercial transactions thatoriginate the transport of electricity. Point tariffs only depend onthe energy injected or taken in each node. Point tariffs may bedesigned to reflect the costs of using the grid. In addition, pointtariffs that are sensitive to location — nodal and zonal tariffs —serve to manage congestion.

Alternatively, there is a transaction-based approach to tariff setting.It consists in setting point-to-point tariffs that depend on thesource and sink of each individual transaction. Contract path anddistance related tariffs are two common examples of this approach.Transaction based tariffs are not in general cost reflective49 and donot serve to manage congestion. However they have been — andstill are — widely used. Table 12 lists the main pricing techniqueswithin each group, which are discussed and compared below.

104

networks === 7

Non-Transaction Based Nodal

Pricing Zonal

(Point Tariffs) Postage Stamp

Transaction Based Contract Path

Pricing Distance-related

(Point-to-point Tariffs) Etc.

Table 12

Pricing of Network Services

49. This is neatly illustrated in the following example. Consider two simultaneous transactions for the same amount ofenergy. In one transaction energy is generated in location A and delivered in location B. In the other transaction energy goesfrom location B to A. Clearly the two transactions cancel out and no transmission services would be needed. However, if thetransmission tariff is transaction based, it will be charged twice.

Short-term Pricing Issues:Managing Congestion

■ Nodal Pricing

Nodal prices equate supply and demand of electricity at each nodeof the transmission grid. Nodal prices are continuously adjustedover time and are set for delivered energy, including both the priceof energy and the price of transmission. The price charged fortransmission is therefore implicit in nodal prices. Nodal pricing isalso known as locational market clearing pricing and as the split themarket approach.The conceptual foundations of nodal pricing aresummarised in Box 3 (Chapter 6).

Until recently, nodal prices were not used to allocate transmissioncapacity and to deal with congestion. Instead, most electricitysystems have relied (and still rely) on other pricing techniquescombined with non-price mechanisms. Nodal prices are now beingused in some US pools and in New Zealand, often in tandem withoperational separation of transmission and generation.

How does nodal pricing match up to the various objectives set outat the beginning of this chapter? Several criteria are relevant:financial sufficiency, efficiency, implementation and competitiveneutrality.

As regards financial sufficiency, in practice nodal prices generaterevenues well below historical cost. In an “optimally planned”system, there should not be such a discrepancy50. However, existinggrid and generation assets are the result of incrementalinvestments over long time periods and under changing technicaland economic conditions.Thus, in practice, nodal prices have to becomplemented with a fixed charge on transmission to collect theadditional revenue.

As regards efficiency, nodal prices reflect the relative scarcity oftransmission capacity at each point of the grid. This providesincentives for efficiency both in the short and in the long-term.

105

networks7

50. Nodal prices would provide the right amount of revenue if (1) generation and transmission assets wereoptimally planned so that, in particular, there would be no excess generation or transmission capacity and thelocation of assets would satisfy a total cost minimisation criteria; and (2) supply were competitive.

With nodal pricing, transmission is relatively expensive at thosenodes in which there is not enough transmission capacity availableto accommodate all scheduled transmission. Higher pricesdecrease demand for electricity, thus resolving congestion (short-term efficiency). Higher prices also provide incentives to invest in interconnections to high price areas (long-term efficiency).Whether the incentives to invest under nodal pricing aresufficiently strong to eliminate all un-economical bottlenecks willbe discussed below.

Finally, there are the issues related to implementation andcompetitive neutrality. Creating a market for transmission servicesraises concerns about the exercise of market power by the ownersof transmission assets and the system operator.The main concernis that the owner(s) of transmission assets may manipulate prices.For instance, holding transmission capacity may artificially createcongestion and thus raise prices. An additional concern is thatsetting nodal prices requires centralising information to calculateprices and to re-dispatch generation in an efficient way. Thus,the system operator has considerable power in shaping marketdecisions, which creates opportunities for monopolistic abuse.

It has been argued that a decentralised implementation of TradableCongestion Contracts (TCC) may provide protection againstmonopolistic abuse. Under this approach,TCCs are sold to marketparticipants through an auction and are then traded in a secondarymarket.TCCs protect market players against changes in the priceof transmission and the secondary trading limits the role of thesystem operator in setting prices. However, it has been shown thatthis approach may exacerbate market distortions when there ismarket power, its performance is not proven and may beinadequate in some realistic situations51.

Nodal prices allow for an efficient management of congestion, butthey are not immune to manipulation when there is market power.

106

networks === 7

51. These criticisms are discussed in Henney (2000), Joskow and Tirole (2000), Hogan (1999), Hogan (2000)and Joskow (1997).

In addition, most often, nodal prices need to be complementedwith other charges in order to raise enough revenue to coverhistorical costs.

■ Zonal Pricing

Zonal pricing is a simplified version of full nodal pricing.The controlarea of the system operator is divided into zones, and prices areset for each zone averaging the cost of congestion of the nodeswithin the zone. It relies on the assumption that congestion tendsto occur in just a few nodes of the grid. Its main appeal is that it iseasier to implement than full nodal pricing.

Other aspects of the performance of zonal prices are similar tonodal prices. Zonal prices also fail to provide enough revenue (thusneeding to be complemented with other charges), impose risks onmarket participants (that can be hedged by means of TCCs) andare subject to manipulation when there is market power.

Does the simplicity of zonal pricing outweigh the presumablyhigher efficiency of nodal pricing? As experience with zonal andnodal pricing grows, the case for nodal pricing is gainingmomentum. Zonal pricing may perform well in somecircumstances, but its apparent simplicity can be misleading. Zonalpricing requires setting mechanisms to deal with intra-zonecongestion in addition to setting prices to deal with inter-zonecongestion, which adds significant complexity. Also, as « know-how » improves, concerns about the complexity of nodal pricingare gradually vanishing. Furthermore, zonal pricing has notperformed well in some instances. It was tried in 1997 in the PJM.Congestion was underpriced, so market participants scheduledmore bilateral transactions than could be accommodated by thegrid. Hence, the system operator had to intervene administrativelyto preserve reliability, constraining choice in the market52.

107

networks7

52. OECD/IEA (1998).

■ Postage Stamp Pricing

A flat rate is set over pre-specified time periods. It gives the rightto inject energy at any node of the grid and to take it at any other.This implies that non-price methods have to be applied to managecongestion whenever it arises. Nevertheless, postage stamp pricinghas the advantage of simplicity and transparency: prices are knownin advance and are easily controlled by the regulator, and themarket for generation is separated from the market fortransmission.This approach is widely applied in EU countries.

Postage stamp pricing, while institutionally simple and transparent,is generally inefficient. Other pricing methods such as nodal pricinghave a clear advantage over postage stamp pricing whenevercongestion problems are significant. However, this can be areasonable approach whenever congestion problems are small. Astrong grid and large reserve capacity in generation are not unusualin many OECD countries. In these countries, congestion may berare, and the benefits of optimal pricing relative to a flat rate maybe small and not compensate for the costs and increasedcomplexity of full nodal pricing. For instance, it has been estimated53

that the losses created by inefficient transmission pricing in the UKare small (some 0.6% of generators’ revenue).This argument couldjustify the use of postage stamp pricing in several EU countries.

■ Transaction-based Approaches

Two common forms of point-to-point tariffs are:

■ Contract Path Pricing: Prices are set for each transmission linein the grid. Each transaction is assigned a “contract path” overthe grid joining the location of the buyer and the seller. Theprice charged to the transaction is the sum of the prices of thetransmission lines crossed by the contract path. While thismethod may seem simple and intuitive, the contract path doesnot reflect the actual flow of electricity over the network or its

108

networks === 7

53. Green (1998).

cost. Thus, contract path pricing does not provide an efficientmanagement of congestion.This method is sometimes used inthe US.

■ Distance-Related Pricing: Prices are set as a function of thedistance between buyer and seller. The method is similar tocontract path pricing and has the same pitfalls. Distance relatedpricing of transmission is applied in Germany.

In general, transaction based approaches are unsatisfactorybecause they result in prices that do not reflect costs and do notserve to manage congestion efficiently. In addition, they may alsohave anti-competitive effects. Their implementation requires thecommunication of information on commercial relationships thatmay be strategically sensitive.They may also favour discriminationagainst some competitors. For instance, distance related pricestend to impose a larger cost on far away and foreign generatorsthat does not necessarily reflect the cost of transportation.

■ Other Methods to Deal with Congestion

In the absence of full nodal pricing, additional mechanisms areneeded to allocate transmission capacity. Capacity can be allocatedon the basis of priority rules that determine a pecking orderamong generators and buyers of electricity. Also, the right of access can be allocated by means of long-term contracts. However,these two mechanisms are generally inefficient and possiblydiscriminatory.

Non-discriminatory mechanisms can also be designed.An exampleof a non-discriminatory mechanism is the so-called “counter-trade” approach in which a parallel generation market — thebalancing market — is run to deal with congestion. The systemoperator directly balances the market making use of the balancingmarket. This is achieved by re-dispatching generation unitscontracted for in the balancing market. The prices of energy andtransmission are those that would have resulted in the absence ofcongestion, and sellers and buyers conduct their transactions as if

109

networks7

there were no congestion. As a result, counter-trade does not giveadequate signals to market players (who ignore transmissionconstraints) and some short-term efficiency losses may occur. Onthe other hand, counter-trade may encourage competition in theelectricity market because, from the players’ point of view, themarket is not segmented by transmission constraints, thus limitingthe scope for the exercise of market power.

Another non-discriminatory approach consists of conducting anauction whenever transport capacity becomes scarce. This mayyield an efficient outcome provided the auction is appropriatelydesigned. It is, however, impractical when congestion arisesfrequently.

Setting Price Levels: Cost of Service VersusIncentive Regulation

Network tariffs are set to provide a certain amount of revenue tothe network service provider. Traditionally, regulated firms havebeen allowed to earn just enough revenue to cover their historicalcosts including a return on investment that corresponds to thecost of capital. This is known as “cost of service” or “rate-of-return” regulation. It relies on the book value of assets, allowingcompanies to recover accounting costs and to earn a “fair” returnon investment. Inflation adjustments and depreciation schedulescan be superimposed. Cost of service regulation is fair, in the sensethat it does not allow the regulated company to make any“extraordinary” economic profits, and is (by definition) financiallysustainable. In addition it provides strong incentives for investmentas prices are adjusted to ensure that investors recover theirinvestments (plus a profit). Cost of service regulation has, however,some well-known drawbacks.This approach provides no incentivesfor cost efficiency and, indeed, it provides incentives to overinvestand to overspend.

110

networks === 7

Incentive regulation is an alternative to cost of service regulationthat aims to provide incentives for cost efficiency. Incentiveregulation allows the regulated firm to retain temporarily some (orall) of the benefits resulting from efficiency improvements. Thisprovides regulated firms with an incentive to reduce costs but itallows prices and revenues to temporarily exceed costs.

Incentive pricing in transmission and distribution (as well as inother regulated industries) is usually implemented by means ofprice caps. Prices are set to cover historical costs, including areturn on investment, minus a given fraction, X, of this cost. Theregulated company is allowed to retain all the additional profits ifcosts are reduced by more than X. However, it has to assume allthe losses if costs are reduced by less than X. Incentive regulationis typically implemented by means of a “RPI-X” formula that allowsyearly price increases of X% points below inflation (so if inflationequals I, transmission prices increase by I-X).This method is appliedto a number of transmission companies.

There are other incentive pricing mechanisms, including:

■ Yardstick and benchmark pricing: Price is set equal to theestimated cost of providing the same service by othercompanies. It has a practical advantage over some othermethods such as price cap regulation where X is very difficultto calculate. This “competition by comparison” approach isintended to provide a benchmark that is not influenced by theregulated company. It is more widely applicable in thedistribution of electricity (and other services like water supply),where several comparable companies operate, than totransmission. However, an international (or interregional)benchmarking of transmission companies can be applied insome circumstances.

■ Sliding scale regulation:This is similar to price cap regulation butthe company is allowed to retain only a fraction of the profitsobtained from efficiency improvements.The fraction of retained

111

networks7

profits decreases with the amount of profits obtained, and theyare typically shared between the regulated company andconsumers54. This method is intended to reduce the profitsmade by firms under price cap regulation which are sometimesdeemed excessive but it has not been actually applied.

Experience with incentive regulation suggests that the potential forcost reductions in transportation activities is large. For instance, inthe UK the introduction of incentives for the National GridCompany to reduce the cost of “uplift” (i.e., ancillary services)resulted in a decrease in costs from £800 million in 1994/95 to£360 million in 1998. Despite its apparently good performance,incentive regulation has been subject to criticism on severalcounts. First, some companies subjected to a price cap have beenhighly profitable suggesting that price caps could have been lower.More frequent regulatory reviews to set price caps and “slidingscale” regulation allow for a more rapid transfer of cost reductionsto consumers at the expense of incentives. Second, incentiveregulation has been criticised for relying too heavily on regulatorydiscretion (e.g. there is some inevitable degree of discretion insetting the “X” factor in a price cap). Regulatory discretion may runagainst the interest of investors if “X” factors are high or, moreplausibly, against the interest of network users if pressure onregulators results in low “X” factors.

In addition to incentives, a related issue is that cost of serviceregulation is based on historical costs. Historical costs reflectexpenditures that were made in the past and that may not beeconomical under current conditions. As a result, when prices arebased on historical cost, they may fail to provide adequate signalsto investors and buyers and sellers of electricity. Ideally, pricesshould be based on the marginal cost of providing the serviceunder current conditions, instead of looking backwards.

112

networks === 7

54. In other words, a sliding scale is a price cap in which the “X” factor grows with profits.

The Institutional Framework: the Role of System Operators in Investment

A key and partly unsettled issue is defining the role of the systemoperator in managing transmission in the long-term, that is, theplanning and implementation of investment in the network.Thereis a broad consensus that the system operator (or otherappropriately designed entity) needs to retain some responsibilityin grid planning and augmentation. Incentives for grid investmentmay be distorted in a number of ways:

■ Market power may reduce incentives to invest (e.g. bottleneckscreate rents for generators).

■ The risk of free riding may discourage investment. For instance,future incremental investments may significantly decrease thereturn on past investments, due to large price changes,discouraging grid augmentation.

■ Incentives for maintenance and replacement of assets may beweak unless appropriate rewards and penalties for security andreliability are designed.

■ Opposition from environmental groups, lengthy administrativeprocedures and other non-economic factors, may imposeadditional costs and delays in the development of transmissionassets.

This suggests that market-driven investment alone could result inunderinvestment in grid expansion and that the system operator(or other appropriately designated party) should retain someresponsibility in grid planning and augmentation. Inappropriate oruntimely grid investments may significantly decrease the efficiencyof the ESI and have a negative impact on the development ofcompetition. The strategic value of the grid in facilitatingcompetition implies that an effective approach to competitionshould prioritise efficient and timely investments in transmissionassets and should establish the appropriate incentives to eliminatetransmission bottlenecks as quickly as possible. Accordingly,

113

networks7

virtually all liberalised systems assign some responsibility fortransmission planning to the system operator55.

Approaches diverge, however, in what other responsibilities areassigned to the system operator. There are two broad approachesto defining his role56:

■ On the one hand, the system operator can be defined as atransmission monopoly that owns the whole transmissionnetwork and takes on the obligation to provide unlimitedtransmission service, that is, the services that are required foreffective system operation.This is the approach taken in mostcompetitive electricity systems in Europe including the UK,Norway, Sweden, Finland and Spain. In this approach, the systemoperator is responsible for planning grid augmentation andmanaging it. In practice, this approach corresponds to theownership separation model.

■ On the other hand, the system operator can be defined as aresidual provider of services that is not the owner of thenetwork and allows market participants to trade transmissionrights and to invest in transmission assets.This is the approachthat dominates thinking in the US. In this approach, the systemoperator may still have a role in planning but does notundertake investment or does it only as a last resort. Inpractice, this approach corresponds to the operationalseparation model.

Appendix

■ Allocating Costs to Users

An additional issue in the pricing of network services is theallocation of charges to transmission users in order to collect theallowed revenue for transmission.These charges are often needed

114

networks === 7

55. Henney (2000).56. Hogan (1999). See chapter 5 for a comparison of these two approaches.

to cover the deficit in revenues that results from fixed costs notincluded in congestion costs. In the absence of nodal pricing, theunder-pricing of congestion also contributes to generate a revenuedeficit.

Costs provide little guidance in allocating these charges to networkusers due to their fixed nature. Other criteria have to be applied.A number of mark-up rules can be used to set prices for networkservices.These rules consist in setting charges proportional to thedemand of each user (uniform pricing), the price-elasticity of eachuser (Ramsey-Boiteaux Pricing), the marginal cost imputed to eachconsumer (Allais Rule), or the foregone profit or “access deficit”originated by each user (Oftel rule). The proportionality factor isadjusted to yield the desired amount of revenues. For this reason,mark-up rules are also known as fully distributed cost pricing.These rules are relatively easy to implement. However, the artificiallinkage between costs and charges may well result in lack ofsustainability, cross subsidies and other distortions.

A number of special rules have been proposed to apply to verticallyintegrated companies that provide network services together withgeneration or end user supply. In this context, the regulation ofprices has to take into account the incentives provided across thevertically related markets (in particular, generation). These rulesinclude:

■ Efficient components pricing (or Baumol-Willig rule): This ruleis designed to promote productive efficiency among accessseekers.The transmission owner can sell access at a price thatnot only recoups his costs, but also compensates him for anyforegone profits from final sales due to the additionalcompetition from access seekers. This rule, if applied inisolation, is the same as allowing unconstrained monopolypricing of access, thus, it has to be applied in conjunction with aprice cap for delivered energy.

■ Global Price Caps: It has been argued that capping the price oftransmission in isolation may distort the pricing of energy by an

115

networks7

integrated transmission company. A global price cap defines abasket of goods, including transmission, sold by the regulatedcompany. Weights are assigned to each good proportional tothe « quantity » sold of each good, and a ceiling is imposed onthe average price of the basket.

116

networks === 7

INSTITUTIONS AND THE POLICY FRAMEWORK

The Need for Regulatory Institutions to Adapt

Regulatory institutions need to adapt to meet the new challengesposed by reform of the ESI:

■ First, in the new environment, regulatory procedures must betransparent and competitively neutral in order to sustain a levelplaying field for competition. This results in new regulatoryprocedures and, often, in the establishment of new regulatoryagencies independent of the companies which they regulate.

■ Second, the introduction of competition implies thatcompetition law has to be applied to the ESI. This requireseither competition authorities and electricity regulators to takeon new roles to enforce competition law.The relationship withcompetition authorities has to be clarified and effectivecommunication channels between electricity regulators andcompetition authorities, if they are not the same institution,have to be built up.This often means that, during the transition,regulatory capabilities have to be reinforced and moreresources have to be engaged in regulatory activities than in thepast. In the longer term, as these needs recede, competitionauthorities may gradually take over electricity regulation,perhaps retaining a specialised regulatory section for electricityand other network utilities.

■ Third, structural obstacles and political resistance to thedevelopment of a competitive market in electricity often resultin regulatory agencies actively promoting pro-competitivereforms. Indeed, competition advocacy by regulators appears tohave contributed significantly to the advancement of reform in

117

institutions and the policy framework8

electricity and in other previously regulated sectors (e.g. airtransport and natural gas).

■ And fourth, electricity markets benefit from a stable or, at least,predictable regulatory framework. The expectation of a stableregulatory framework may be favoured by independentregulatory agencies within government which are less subjectto political change than other parts of government57.

Regulatory Independence

A crucial issue in the regulation of any industry is theindependence of the regulator.The basic principle is that regulatorshave to be independent from the regulated. Otherwise, conflicts ofinterest are unavoidable, and regulation is bound to deteriorate.Careful design of regulatory institutions is needed to ensureeffective independence of the regulator from the regulated entities.

Independence from government and political actors may also bebeneficial. It helps to ensure stability of regulatory policies, to avoidthe use of electricity policies to achieve general policy objectives(e.g. more revenues from taxation or lower inflation from lowertariffs) and, generally, to protect investors and utilities from short-term political pressures that may undermine the stability of theregime.The importance of political independence for an adequateregulatory performance is likely to depend on a number of countryspecific factors. The crucial issue is to what extent politicalinterference is a real threat.This is influenced by the institutionaldesign of each country. For instance, the role of courts in reviewingregulatory decisions, which is critical in this regard, varies acrosscountries.

Political independence of the regulator is particularly importantwhenever there is public ownership of electric utilities. In this case,

118

institutions and the policy framework === 8

57. This was generally the case of regulatory agencies in the US before liberalisation. However, to the extent thatregulators also act as advocates of pro-competitive reforms, the impact of their actions on the expectations ofmarket players is ambiguous.

the government simultaneously faces responsibilities as owner andas regulator. Instituting a politically independent regulatory bodyavoids potential conflicts of interest between these two areas ofresponsibility.

The independence of the regulator needs to be differentiated fromlack of accountability. Regulatory agencies, like any other publicbody, must be held accountable for their actions and be subject toadequate efficiency controls. Regulatory agencies built on theprinciples of independence (from the regulated) and onaccountability have the highest potential to deal effectively with the new regulatory challenges. However, combining accountabilityand independence is a difficult task. A review of the regulatorystructure must accompany regulatory reform since regulatoryinstitutions designed in the past to deal with other issues may notsatisfy these general principles.

A Review of Institutional Approaches58

Regulatory institutions do not follow a clear pattern acrosscountries. Electricity regulatory agencies that are independent, tosome extent, from other parts of government have existed forover half a decade in the US: the Federal Energy RegulatoryCommission (FERC) at the federal level deals with wholesaletransactions while the different state Public Utility Commissions(PUCs) deal with retail transactions. Many other electricityregulatory agencies have been created more recently, often at thetime of the restructuring of the ESI in Australia, Belgium, Canada,Finland, France, Hungary, Ireland, the Netherlands, Norway,Portugal, Spain, Sweden and the UK.

The division of jurisdictional powers among government, thecourts, the general competition authorities, the national regulatorand, in federal countries, the state regulators largely varies from

119


58. A detailed analysis of regulatory institutions can be found in the forthcoming IEA book RegulatoryInstitutions in Liberalised Electricity Markets.

country to country. In most countries, the scope of the activities ofthe regulatory agency is limited to specific aspects of the industry.Many regulatory agencies have powers to set tariffs, to grantlicences or authorisations, to monitor the regulated companies andto act as arbitrators between private parties. However, in somecountries, the scope of the regulatory agency’s activities isrelatively large, covering most aspects of the specific industryregulation59.

Regulatory agencies are generally set up to be independent fromthe regulated parties and to maintain an arm’s length relationshipwith the political authorities. The regulatory agencies also haveattributes of institutional autonomy, usually including ear-markedfunding and exemption from civil service salary rules60. Funding isoften obtained from levies on regulated firms which, for controlpurposes, is sometimes administered through the general budget(e.g. FERC61).

In most countries, competition agencies are not specifically incharge of ESI regulation, the exception being Australia, where thecompetition authority is also the national electricity regulator. Inaddition, advocacy of competition, the promotion of competitionand defence of pro-competition reforms, is a significant activity fora number of electricity regulators.

The Important Role of Competition Policy

The role of general competition law has to be clarified at the timecompetition starts in the ESI. An active role of competitionauthorities is needed. Oligopolistic conditions prevailing inelectricity generation in many countries require intense

120


59. Regulation which is not industry specific may be handled by a different agency (e.g. the EnvironmentalProtection Agency in the US).60. Smith (1996).61. Specifically, FERC collects revenue for the US Treasury through fees on the industry, but it is actually fundedthrough annual appropriations from Congress.The fees are equal to the appropriated budget.This arrangementallows Congress to retain oversight and budgetary approval while passing costs on to the users through fees.

monitoring of competitive behaviour. Likewise, the development ofa formerly non-existent market in supply to end users calls for acareful scrutiny of anti-competitive practices.

In the past, competition authorities have not been particularlyconcerned with the ESI.Thus, it is important that enough resourcesbe timely assigned to this new task. Also, since detailedunderstanding of the operation of the ESI is needed, close co-operation of competition authorities with electricity regulators isadvisable whenever they are two different institutions.

There are two areas in which competition policy may have a keyrole in the development of competition in electricity markets:merger policy and subsidies. First, the stance taken by competitionauthorities towards mergers and acquisitions is critical in manycountries in which the electricity market is initially highlyconcentrated. Entry and geographical expansion of marketboundaries, which can also mitigate market power, occur onlyslowly in electricity markets. Additionally, merger and acquisitionactivity may seek to reaggregate functions of the ESI, such asgeneration and distribution, that reform has sought to disaggregatein order to promote competition.An active merger control policymay be the only effective remedy against market power in anumber of situations.

Experience in the US and elsewhere shows that merger policiescan be tailored and adjusted to address concerns about marketpower without generally blocking mergers themselves. Identifyingappropriate remedial measures to mitigate market power requiresa careful analysis of markets to identify the geographical areas inwhich market power may be exerted and the particular assets thatconfer market power. Standard antitrust analysis based onconcentration measures is likely to underestimate the potential forthe exercise of market power62.

121


62. See Borenstein, Bushnell and Knittel (1998).

Second, competition authorities have a key role in ensuring thatsubsidies do not distort competition in electricity markets.Subsidies have been common in monopolised electricity markets,including aid to other industries (coal), to specific generationtechnologies (nuclear and some renewables), and to some end usergroups (energy intensive industries). The introduction ofcompetition into electricity markets increases pressure todismantle subsidies but does not necessarily eliminate them.Subsidies may be particularly damaging to competition ininternational electricity markets since differences among nationalpolicies may provide an unfair competitive advantage to someelectricity companies.

Specific tools and expertise for the analysis of competition inelectricity markets are being quickly developed, particularly in theUS where antitrust legislation is most widely applied. While thebasic concepts (e.g. market power) and remedies (e.g. divestituresand facilitation of entry) of antitrust analysis remain the same,market definition and the specific factors that are likely to facilitatethe exercise of market power (or to make it more difficult) in theESI require a refinement of the concepts63.

First, the relevant geographical and product electricity market forantitrust analysis has to be defined with reference to demand andsupply conditions that change over time. For instance, the samegeographical area may constitute just one market in periods of lowdemand but may be split into two or more markets in periods ofhigh demand if transmission constraints impede trade betweenzones.As a result, there may be different markets for different timeperiods. In addition, wholesale and retail transactions generallywould not be in the same market because the opportunities ofbuyers of electricity to switch from one to the other are limited.And, for some purposes, long, medium, and short-termtransactions also define different markets.

122


63. See Frankena and Owen (1994) and Binz and Frankena (1999) for a detailed analysis of antitrust in powermarkets.

Second, the assessment of market power in electricity markets hasto take into account a number of specific factors including the wayin which competition takes place (i.e. how prices are set given aparticular vertical or horizontal structure) and the conditions ofentry into the market. Antitrust analysis often relies onconcentration indices, such as the Herfindalh-Hirschman Index,built from market shares to assess how likely it is that marketpower will result in uncompetitive prices.This approach is seen tobe too broad to make an accurate assessment in the ESI.Increasingly, antitrust analysis of electricity markets relies on fullstructural models of the industry that include an explicitdescription of the vertical and horizontal structures and specificassumptions about the behaviour of firms. As regards entryconditions, antitrust analysis needs to take into account therelatively long (even if decreasing) time span from initial planning tooperation of generation and transmission assets.

Developing a Framework for InternationalTrade

International electricity trade is increasing in many regions of theworld, including the EU and North America. In these regions theESI is rapidly evolving from a set of national markets to become amuch broader regional market.A similar process is taking place infederal countries such as the US and Australia, where moreintegrated national markets are growing out of previously separatestate markets. The development of regional electricity marketsbrings some important benefits. In addition to the direct benefitsof trade resulting from lower overall costs, regional competitionmay compensate for high concentration in domestic markets andencourage the competitiveness of national electricity firms.

However, lack of harmonisation among national regulations mayresult in barriers to trade. First, if some countries are more openthan others within a common trade area, there may be reciprocityconcerns that could make international electricity trade moredifficult. Second, nationally set environmental standards (e.g. post-

123


Kyoto commitments to reduce CO2 emissions) may not beeffective once international electricity trade is engaged. Third,differences in taxation may also distort and discourage trade. Aprominent illustration of how lack of harmonisation may result inmarket fragmentation is provided by the EU internal electricitymarket64.

Extensive international co-operation to deal with these issues isneeded.There are signs that it is increasing.The efforts made in thecontext of the EU Electricity Directive to set common rules, the“flexibility” mechanisms included in the Kyoto Protocol and theincipient analysis of trade of energy services in the context of theGeneral Agreement on Trade and Services (GATS) are examples ofincreasing international co-operation.65 However, more remains tobe done.

A difficult issue that needs to be addressed in the development ofinternational electricity markets is the management oftransmission. In practice, the development of internationalelectricity trade depends on market players having access tointernational interconnectors at a reasonable price. Pricing andallocating international transmission capacity raises similar issuesto those raised by domestic transmission. First, pricing has to avoid the “pancaking effect” of transmitting power across variouselectricity systems, meaning the accumulation of two or morenational transmission charges when this accumulation of charges isnot justified by the cost of the transmission services provided.Second, pre-existing contracts and international agreementsbetween utilities and system operators may conflict with anefficient allocation of transmission capacity.The development of theEU internal electricity market provides an example of these twoproblems66 and of the need for common rules to overcome them.

124


64. See the Second Harmonisation Report, EU Commission (1999).65. See WTO (1999).66. See EU Commission (1998), Comision Nacional de la Energia (1999) and European Transmission SystemOperators (1999).

The Relationship with Natural Gas

Developments in the natural gas market also have to be taken intoconsideration as gas and electricity markets are increasinglyinterrelated.The opening of gas markets and the related changes ingas prices have a potentially large effect on the ESI. Gas prices havea short-term impact on the generation mix that is dispatched. Inthe long-term, investment decisions in generating capacity are alsoaffected.The fuel inputs to the generation of electricity will resultfrom the combined action of changes in both gas and electricitymarkets. In addition, gas companies are actual, or potential,competitors in the electricity market. A wider energy sectorperspective is important in assessing the institutional and otherchanges that are taking place in the ESI.

A convergence of gas and electricity regulation is already apparentin a number of countries. For instance, the regulation of gas andelectricity is assigned to the same regulatory agencies in manycountries.

The Broader Policy Framework:Security of Supply

This book is primarily concerned with the introduction ofcompetition to stimulate greater economic efficiency in the ESI.However, most governments consider they have a strategicobligation to ensure that the ESI is reliable in the short, as well asin the long-term; that environmental objectives are met; and oftenalso that other objectives such as social or regional equity are met.How do these wider considerations fit into a reform processwhich is mainly driven by the objective of improving efficiency?

An important policy priority is to sustain security of supply in itsthree dimensions of short-term system reliability, long-termsystem reliability through adequate investment, and diversity andsecurity of fuel inputs used for power generation. In the past,

125


governments have often promoted security of supply throughdirect intervention. Under competition, security has to bepromoted with more market compatible tools.

■ Short-term System Reliability

There is a need to develop effective rules to facilitate co-ordinationand to ensure the reliability of electricity systems undercompetition. Several objectives have to be considered in designingthese rules, including economic efficiency, minimal commercialinterference and reliability. There is no tailored set of rules forsystem operation under competition. Rules are now beingdeveloped and tested in many countries and it is not yet clearwhich approaches are more effective. From a policy perspective,the most important issue is to assign responsibilities clearly,including an explicit definition of the obligations of each agent andthe financial consequences of failing to meet these obligations.Governments need to ensure that this happens.

■ Long-term System Reliability

In the new ESI, investment in generation is governed by marketrules. In this context, prices signal the relative scarcity ofgeneration capacity, and investment is driven by prices.This marketprocess should result in adequate investment levels and, inparticular, eliminate the incentives to overinvest that arise in aregulated environment. However, it is sometimes suggested thatcompetition may have a negative impact on investment ingeneration assets. First, there is the argument that competition maybe “too effective”, leading to unsustainably low prices that may, inturn, temporarily discourage investment and reduce security ofsupply. However, the real problem has often been the opposite:insufficient rivalry rather than cut-throat competition. Second, ithas been argued that investment may follow a cyclical pattern ifinvestors are myopic, but investors can be expected to take intoconsideration returns over the entire life of the investment.Third,there are concerns that under competition there may be a

126


shortage of investments in peak load capacity, because it will tendto be remunerated randomly and infrequently in a competitivemarket.This argument has already been explored (see Chapter 6on capacity payments).

Even if such concerns were relevant, there are a number of elementsin the ESI that may compensate for any temporary imbalances, thussuggesting that investment in generation capacity should not be anissue under competition. The development of InformationTechnology increasingly allows demand to be responsive to pricechanges without ultimately affecting reliability; and the shortening oflead and construction times for some generation technologiesmeans that supply can react relatively quickly to maintain reliability.

To summarise, the level of investment in generating capacity shouldnot be a concern in a competitive electricity market. However, thetransition to effective competition could create some problems.First, there could be higher-than-normal regulatory risk that maydelay or make it more costly to finance investments. Experience todate (see Chapter 3) indicates that regulatory risk is not merely anacademic concern. Second, inadequate transmission pricing maydistort generators’ decisions. Third, insufficient demandparticipation may result in prices that do not provide the correctsignals for investment.

Adequate investment in transmission and distribution is alsoessential to sustain reliability. Investment can be open to anyinterested parties or, alternatively, responsibility for investment maylie with a single transmission company. In both cases, transmissionprices play a key role in guiding investment decisions. Cost reflective(“nodal”) prices can provide adequate signals for investment.However, transmission networks have some special characteristicsthat may result in investment distortions.There is a broad consensusthat governments (or the appropriate regulated companies) shouldretain at least a residual role in investment in electricity transmission,including monitoring investment developments.This is in addition tothe ongoing role of governments in providing an adequate regulatoryframework for investment.

127


■ Diversity and Security of Fuel Inputs

It is not clear which fuel and technology choices will develop undercompetition. A competitive market provides incentives forinvestors to choose the least cost alternatives. For instance, givencurrent conditions, fossil fuels such as gas often provide electricityat least cost.This, of course, may conflict with national policies thatgive preference to other fuels (e.g. nuclear promotion in somecountries and closing of nuclear plants in others) or that wouldfavour a different fuel mix (e.g. more domestic coal and less naturalgas in electricity generation).

Governments can, to some extent, continue to implement fuelpolicies in a competitive market. For instance, governments canrestrict fuel choices for new entrants or can establish subsidies forgeneration based on certain fuels. However, this approach may havea significant impact on market performance, as it distorts entry andinvestment decisions. Ultimately, regulations on entry are likely toconflict with the efficient operation of the market, and reduce theefficiency gains that were the primary objective of reform.

The Broader Policy Framework:the Environment

Protection of the environment is a growing policy priority. Thisobjective includes a broad array of issues ranging from post KyotoCO2 emissions reduction targets, and other air and waterpollution control, to nuclear safety and radioactive waste disposal.Monitoring and enforcement of established environmentalstandards remains a public policy issue in a competitive market.Indeed, once environmental standards are set, strict enforcement isimportant also from a competition policy perspective to ensurethat no agent unfairly enjoys a cost advantage from non-compliance.

128


Policy makers face the challenge of designing least cost policies topursue these objectives, that is, policies that reduce or eliminateenvironmental externalities (or promote social objectives) whileminimising market distortions. In general, this means thatenvironmental objectives have to be implemented withcompetitively neutral, non-discriminatory, mechanisms and that theuse of other mechanisms must be minimised. A broad set ofeconomic incentives to promote environmental objectives isavailable. Tools like tradable emission permits or emission taxeshave the potential of improving environmental performance whileminimising market distortions. Other market-friendly policy toolsthat have been used to pursue environmental goals are renewableportfolio standards, enhanced public research and development(R&D) budgets, and non-bypassable grid charges established tofinance public benefits.

Environmental protection regulations already have a large impacton the ESI, and this impact is likely to increase in the futurethroughout OECD countries. Regulations include specific actionson both the demand and the supply side of the electricity marketas well as general economic instruments that act simultaneously onboth sides of the market and on other markets.

On the supply side, there are requirements on generating plants toreduce emissions of pollutants such as sulphur and carbon dioxide.These take the form of standards and quotas that may be assignedeither to individual generation units or globally to the industry.

Support programmes aim to increase the use of environment-friendly technologies, particularly renewables, and to increase theefficiency of power generation, specially by promoting co-generation. These measures include investment subsidies, fiscalincentives, green pricing, guaranteed markets and/or subsidisedprices, tradable green certificates markets (i.e., consumers have anobligation to buy “green certificates” from “green generators” inproportion to their electricity consumption), government-supported R&D, standards and obligations to adopt certain

129


technologies. Such programmes have already significantly increasedthe share of renewables and CHP in power generation.

On the demand side, policies include demand side managementprogrammes, aimed at improving energy efficiency at theconsumption point, government-supported R&D and consumereducation programmes.

The difficulty with current approaches is that they do not optimisethe economic efficiency of meeting environmental objectives in aliberalised ESI, as they have a very significant potential to distortmarket decisions. Hence, there is growing emphasis on so-calledeconomic (or market compatible) instruments to achieveenvironmental objectives.

Economic instruments, including taxes and tradable quotas, arethose that result in electricity producers receiving the same priceregardless of its source, and all emitters (or any other targetedexternality) paying the same price regardless of source. The firstcondition is necessary for an efficient allocation of electricity: sinceall electricity renders the same service regardless of its source,price should not depend on the source. The second condition isnecessary for an efficient — least cost — allocation of abatementcosts among polluters.

A tax on the externality — for instance, on each tonne of carbondioxide discharged into the atmosphere — is designed to makepolluters internalise the social cost of emissions. There are twomain advantages of environmental taxation compared with otheroptions. First, environmental taxes are efficient, that is, they can“price away” the externality while minimising distortions on otheraspects of economic activity. Second, the cost to tax payers isknown in advance, while the cost of regulatory measures is onlyknown after they have been implemented.Assigning environmentaltaxes, however, requires governments to develop measurementand collection mechanisms, and so the implementation of anenvironmental tax is not costless. Environmental taxation is beingconsidered in the EU and, individually, in some EU member

130


countries (e.g. UK and Denmark) as a tool to meet CO2 reductioncommitments. In practice, the EU approach may involve a tax leviedon all energy, combined with a series of exemptions and refundslinked to the contribution to achieve emission reductionobjectives.

An alternative to taxation is to set pollution quotas limiting theamount of pollutants that can be generated. Pollution quotas haveone key advantage over taxation. Unlike taxation, a quota impliesthat the amount of pollution that will be generated is known inadvance. However, the cost imposed on the generators of pollutionis not known at the time of setting the quota. Conversely taxationimplies that the cost to taxpayers is known in advance but theamount of pollution that will be generated is uncertain.

A pollution quota, like a tax, requires governments to set upmechanisms to measure pollutants and to enforce the quotathrough penalties for non-compliance which, implicitly, will act as aceiling for the actual unit cost of the quota. In addition, an efficientquota system requires the development of a trading mechanismamong polluters. For instance, governments may establish the totalamount of carbon dioxide emissions allowed in their country in agiven year. Setting limits for each polluter within the country wouldrequire very detailed and disperse information which is notavailable to governments. An aggregate approach, such as forcing allpolluters to cut their emissions of carbon dioxide by X %, wouldnot be efficient because the cost of reducing emissions is lower forsome industries and technologies and higher for others.To assignthe pollution quota to the polluters efficiently, that pollution shouldbe reduced by industries with the lowest abatement technologies.Governments (or any other single party) are not in a good positionto assess differences in costs across industries and firms.

One approach to minimise the total cost of a pollution quota is tocreate a market for emissions. This amounts to establishing anaggregate emissions limit, distributing to individual sources anumber of emission permits equal to this limit following somecriteria, and allowing the permits to be traded. Emissions trading is

131


efficient because firms with lower emissions abatement costs mayprefer to abate more and to sell their allowances to firms withhigher emission abatement costs.The price of the emission permitsconveys valuable information to traders on the cost of reducingemissions and, in this way, trade lowers the overall cost ofcompliance with the quota. Experience with this approach is stilllimited. It has been tested with some success under the US acidrain programme to limit sulphur dioxide emissions. Otherprogrammes developed in the US covering a range of air pollutantsand water effluents have been less successful.There is virtually noexperience with tradable permits outside the US.

Tradable permits for carbon dioxide « equivalents », includingcarbon dioxide itself and five other greenhouse gases which aremeasured in terms of the former, have been proposed in thecontext of the Kyoto Protocol.At this stage, the implementation oftradable carbon permits bears two main difficulties. One is theallocation of the pollution quota within each country. If countriesadopt different allocation systems, the same industries will beardifferent costs in different countries and this may affectinternational competitiveness. The second difficulty is that offinding a consistent set of national and international enforcementsystems ensuring, for instance, a common penalty level. Theperformance of emissions trading mechanisms depends on anumber of factors, including the size of transaction costs and the« liquidity » of the emission permits.

Technology support measures and, to a lesser extent, emissioncontrols currently dominate the international scene for meetingenvironmental objectives. Economic instruments are not widelyused.This picture may change if environmental policies in responseto the threat of climate change become important. The KyotoProtocol commits the Annex I countries, including IEA countriesand Economies in Transition, to cut their greenhouse gasemissions by about 5% relative to their 1990 levels. Reductiontargets have to be met from 2008 to 2012 and must cover fiveother greenhouse gases in addition to carbon dioxide. Specific

132


reduction targets vary from country to country, ranging from an8% reduction for most countries to increases of up to 8% forother countries. The Protocol will enter into force 90 days after55 countries, accounting for 55% of total carbon dioxideemissions of Annex I countries, have ratified it. When this couldhappen is subject to considerable uncertainty as ratification of theaccord by some countries, notably the US, is still uncertain.Overall, while there is an expectation that greenhouse gasemission control policies will develop, there is considerableuncertainty as to which particular policies will be applied toachieve the emissions reduction goal67.

From the point of view of competition and market performance,economic instruments have a number of advantages over otherpolicy instruments. In particular, economic instruments addressexternalities without distorting competition or inducinginefficiencies, but they are not yet widely applied in practice. Anumber of factors, including political constraints, macroeconomic(e.g. fiscal policy) considerations, lack of internationalharmonisation, and concerns about the competitiveness ofparticular industries, have resulted in environmental policies thatare largely based on other, less efficient instruments.

In principle, taxes and tradable quotas are equivalent in that bothhave the potential to resolve the externality without inducingdistortions in other parts of the market. In practice, however, thereare differences because the cost of reducing emissions is highlyuncertain. Thus, the actual cost of a tradable quota may differsignificantly from the expected cost.The actual cost of taxation, onthe other hand, is less subject to variation. This provides an

133


67. The Protocol includes or allows a number of mechanisms aimed at minimising the costs of implementationby allowing flexibility in the timing, the place and the subjects of reduction targets. These mechanisms allowflexibility within Annex I countries by means of joint implementation, international trading of GHG and bubbling,i.e. joint commitments by several countries.These mechanisms also allow flexibility between Annex I countries andthe rest of the world under the Clean Development Mechanism. In addition, the Protocol sets an overall target tobe reached by the end of the agreement period (2008-2012) but does not constrain the path of emissionsreductions over time.

argument to choose taxes over quotas as the preferredinstruments of environmental policy68.

The impact of other policy instruments on competition and marketperformance differs. Some policy instruments, such as adequatelytargeted R&D subsidies, are competitively neutral and thereforeneed not conflict with the introduction of competition intoelectricity markets. Other instruments, to different degrees, distortthe market by creating separated markets for green and non-greenelectricity, subsidising some competitors, or forcing (possibly non-cost effective) technological choices.

A fundamental issue in assessing the potential impact of these “noneconomic” instruments is whether they are cost effective inmeeting their goals. Consider, for instance, technology supportpolicies aimed to increase the share of renewables in powergeneration.A number of alternative measures can achieve the goalthat X% of all energy be produced from renewable energy sources.Possible measures include a guaranteed price for green energy, areserved market for green energy (i.e., some X% of all energy hasto be purchased from green generators) and a tradable greencertificates market.The important issue is the cost effectiveness ofthese alternatives.

Market-based approaches can be expected to perform better thanthose based on administrative intervention because they allowmarket players to search for the most cost-effective alternative. Inaddition, larger and more flexible markets can be expected toperform better than more restricted markets. Thus, a greencertificates market can be expected to be more cost effective thanthe other two options. In particular, market-based approaches havea significant potential when environmental policies commitsubstantial resources and the underlying market is large andpotentially efficient. On the other hand, setting a guaranteed pricehas lower start-up costs, and may be an easier and more

134


68. Koops (1999).

straightforward option when the underlying market is small. Thissuggests that, while administrative approaches have dominated thefirst steps of environmental policy, market-based approaches willbe needed if environmental policies are escalated.

The Broader Policy Framework: Social and Regional Objectives

To varying degrees, countries have also pursued social and/orregional objectives in connection with the ESI and these objectivesmay continue to be important. Governments may, however, need toreview and clarify what social and regional objectives they want topursue. Policy priorities of the past may no longer be important.For instance, uniform electricity tariffs may no longer be justified ifbroad geographical cohesion ceases to be an important policy goalbecause it is already achieved. Social objectives, including universalservice, support to disadvantaged consumers or equalisation oftariffs can, in principle, be promoted in a market context providedthey are made explicit and they are financed in ways that do notdistort competition. Achieving these objectives, nonetheless,generally entails distorting electricity prices and therefore there isa limit to what electricity markets can contribute to socialobjectives without suffering significant loss of economicperformance. In the long run, general policies, rather than sectoralelectricity policies are more efficient in achieving social objectives.

In defining social objectives, policy makers should keep in mind thatcompetition itself brings some significant social benefits. Forinstance, low prices are an important step forward in achievinguniversal service; and more product and price differentiation havethe potential of allowing a better match between special consumergroups and the service they need.

135


CONCLUSIONS AND OUTLOOK

Consumer Choice is the Key to Reform

A fundamental pillar of effective reform in the ESI is consumerchoice, that is to say, giving all consumers the ability to choose theirelectricity supplier. Consumer choice disciplines market players,because dissatisfied consumers can switch supplier, and itencourages innovation. As in any other market, consumer choice inthe ESI is fundamental to achieving both static and dynamicefficiency, and it is difficult to envisage real competition without it.There are virtually no examples of markets in which competitiondoes not depend on consumer choice.

Competition in wholesale markets has resulted in significantimprovements in the productivity and internal efficiency of electricitycompanies but translating these benefits to consumers has proven tobe a crucial challenge for electricity market reform. Experience hasshown that consumers benefit from reform only if the otherelements of the supply chain transmit the benefits of competition.This means that there is an increasing emphasis on the regulation oftransmission and distribution to ensure non-discriminatory accessconditions and cost-oriented tariffs, and that competition in end-user supply is increasingly seen as a means of translating toconsumers efficiency gains obtained upstream in the supply chain.

Challenges

A major challenge which early reformers have had to tackle is thatlifting barriers to entry and changing the rules is not enough:competition requires a sufficient number of competitors. If supplyis concentrated in a few firms, competition generally fails todevelop and prices may remain persistently above theircompetitive levels. In the electricity sector, concentration has often

137

conclusions and outlook9

resulted in prices that are too high69. The UK experience, forinstance, showed that the performance of a duopoly was notsatisfactory and that further measures were needed to reduceconcentration after privatisation70.

High concentration is a common feature of the ESI. In the past,power supply has tended to be concentrated in a small number offirms. Because of the lack of competition, seller concentration wasnot considered a problem. Liberalisation policies, aimed at marketentry and prices may not be enough, at least in the short-term, toestablish effective competition. Restructuring policies may also beneeded to correct inappropriate market structures.These policieshave been developed in most liberalising countries either at thetime of liberalisation or later, when market performance hasindicated the need for a less concentrated structure.

However, it is also important to note that “workable” rather than“perfect” competition is the goal. Workable competition mayemerge even if suppliers are not “atomised”. Most experts agreethat four or five comparable power generating firms, under theappropriate circumstances, may yield a satisfactory marketperformance or, at least, one that improves upon the initialsituation. If this view is correct, relatively limited restructuringpolicies may suffice to obtain workable competitive markets. Anadequate market structure in other parts of the supply chain is alsoessential to develop workable competition. In particular, effectiveconsumer choice requires the development of a critical mass ofretail suppliers (who may not be generators).

Another challenge is to develop techniques and approaches tomeet objectives for the ESI, including efficiency, security of supply,environmental protection and social goals, in a market compatibleway. Development of regulation compatible with undistorted

138

conclusions and outlook === 9

69. This was not totally clear until the first competitive pool started operation in the UK. It is possible that intenseprice competition may develop even if the number of competitors is low (“price wars” among duopolistic aircarriers, for instance, are not uncommon).70. A survey of studies on market power in electricity markets can be found in Kahn (1998).

market performance is a key challenge for most regulatory systemsas the required know-how and procedures involved in developingmarket-oriented regulation are radically different from thoseneeded without competition. In a competitive environment,economic factors tend to outweigh technical factors in definingfeasible regulations; regulators have to be independent from theregulated entities, and regulatory procedures have to be explicitand transparent.All these elements force some important changesin the management of regulation, typically leading to areinforcement of the resources devoted to regulation and, moreimportant, to more independent and transparent regulatoryinstitutions. One particular challenge is to develop effectiveregulatory approaches to transmission systems, which presentdifficult characteristics and which remain an essential input toelectricity supply.

Security of supply, and environmental and social objectives have tobe pursued with suitable instruments that include taxes to priceexternalities, non-discriminatory subsidies for the development oflong-term R&D projects, market mechanisms (tradeable emissionpermits), performance standards and some general obligations(obligations on suppliers to contract with renewable fuel powerproducers). Command and control mechanisms are not the bestapproach in this new context. Also, social objectives wheneverpursued have to explicitly and transparently regulated; for instance,obligations to supply and codes of practice for some categories ofcaptive end users such as the elderly and disabled.

This has proved to be a difficult task. Transition and broadereconomic policy issues often slow the reform of the regulatoryframework. As an example, the taxation of CO2 emissions raisesissues about the differential impact on energy-intensive industries,the competitiveness of national economies and the overall designof fiscal policy.As a result, taxation of emissions is still rare in theinternational arena.

The environmental challenge is perhaps the biggest.The last thirtyyears have seen rapidly growing concern about the environmental

139


effects of energy production, including acid rain, nuclear wastedisposal and, more recently, the issue of climate change.Environmental protection is consequently a top priority forgovernments.

Further Change Can be Expected

Electricity market reform is a moving target. Many studies on thesubject written only a few years ago are already outdated. One keydevelopment is the market itself. The ESI is rapidly expanding itsboundaries. Generation is increasingly integrated with gas and oilcompanies. Vertical integration allows generators to hedge risksrelated to the availability of fuel. Simultaneously, integrationreduces risks for gas and oil companies by securing a relativelystable demand for their products.Two factors work in combinationto increase vertical integration. First, gas-fired generation isgrowing around the world. Second, the liberalisation of generationfacilitates vertical integration by removing regulatory barriers toentry into generation.

At the same time, distribution and end user supply activities innetwork industries, such as gas, electricity and some areas oftelecommunications or water, are likely to become moreintegrated across industries. The reason is that there areeconomies of scope in end user activities and, to a more limitedextent, in the distribution of services. Again, the opening tocompetition facilitates this trend.

The geographical boundaries of the ESI, once coincident withnational or state boundaries, are also changing. Increasingly,electricity systems are becoming integrated within regionalmarkets. Liberalisation is also opening the way for significant directinvestment by foreign companies in national markets. Countries inwhich electricity companies are owned by foreign companies orinvestors include the US, UK, Australia, New Zealand and manynon-OECD countries.

The implication for policy of this expanding industry boundary isthat regulation will have to be managed more and more at a multi-

140

conclusions and outlook === 9

sectoral and multinational level. This evolution towards greaterdependence on general regulation, competition law, and commonrules for international trade is a process that has already occurredin many other industries opened to competition.

Technology is also changing.The emergence of “e-commerce” willhave an important impact on transactions and market behaviour in the ESI. Internet-based electronic commerce providesopportunities for companies to develop by expanding services,providing new choices, streamlining processes and reducing costs.These new processes are also creating new business models thatwill continue to change the shape of the ESI and related industries.

Another, even more important, technological development on thehorizon is the widespread adoption of non-centralised (self-)dispatch. Autoproduction and distributed generation are rapidlygrowing in many countries due to the development of efficientsmall-scale generation. Environmentally, these developments arevery positive to the extent that they make use of renewablesources of energy.

Distributed generation and autoproduction are both substitutesfor electricity transportation services. This has the effect ofweakening the effectiveness of natural monopoly regulation of thenetwork. If electricity is generated closer to where it is consumed,relatively fewer transmission and distribution services will beneeded, with large potential implications for network regulation.For instance, if transportation services are not used by allconsumers, the effectiveness of transportation charges as a meansof financing stranded and fixed costs will be reduced. In due course,if new generation technologies become a profitable substitute fortransportation, transportation will cease to be a natural monopoly.Such a development would have a crucial impact on the futureeconomics, market structure and regulation of the ESI.

141


BIBLIOGRAPHYARMSTRONG, M. and DOYLE, C., 1995.“The Economics of AccessPricing”. Regulation and Incentives in Telecommunications.Competition and Consumer Policy Division. OECD.

Australian Competition and Consumer Commission, 1999.“Guidelines on Undertakings”. ACCC web page, 15 August 1999.

BERGARA, M. and SPILLER, P., 1996. “The Introduction of DirectAccess in New Zealand’s Electricity Market”.Working Paper of theUniversity of California Energy Institute (Berkeley). PWP-043.

BINZ, R. and FRANKENA, M., 1999. “Addressing Market Power”.Policy Papers of the Competition Policy Institute.Washington, DC.

BOND, J., 1998. “The Asian Crisis and Power Reform in EmergingEconomies”. Paper presented to Montreux Energy Roundtable, 19May 1998.

BORENSTEIN, S., BUSHNELL, J. and STOFT, S., 1996. “TheCompetitive Effects of Transmission Capacity in a DeregulatedElectricity Industry”. University of California Energy InstituteWorking Paper. PWP-040.

BORENSTEIN, S. and BUSHNELL, J., 1996.“An Empirical Analysis ofthe Potential for Market Power in California’s Electricity Market”.University of California Energy Institute Working Paper. PWP-044r.

BORENSTEIN, S. and J. BUSHNELl, 1998. “Market Power inElectricity Markets : Beyond Concentration Measures”. Universityof California Energy Institute Working Paper PWP-059.

BORENSTEIN, S. and BUSHNELL, J., 2000. “ElectricityRestructuring : Deregulation or Reregulation ?” University ofCalifornia Energy Institute Working Paper PWP-074.

BORENSTEIN, S., BUSHNELL, J. and WOLAK, F., 2000.“DiagnosingMarket Power in California’s Deregulated Wholesale ElectricityMarket.” University of California Energy Institute Working PaperPWP-064r.

143

bibliography

BUSHNELL, J. and STOFT, S., 1996. “Transmission and GenerationInvestment in a Competitive Electric Power Industry”. Universityof California Energy Institute Working Paper. PWP-030.

BUSHNELL, J., 1999. “Transmission Rights and Market Power”.University of California Energy Institute Working Paper. PWP-062.

Cambridge Energy Research Associates, 1998. “ElectricityLiberalization in Europe”. Private Report Series, January 1998.

COSTELLO, K., 1998. “Fair Competition in Retail ElectricityMarkets”. Mimeo. Unpublished Working paper. National RegulatoryResearch Institute. Ohio State University. June 1998.

CPUC, 2000. “Report to the Governor on the electricityconditions facing California”. California Public UtilitiesCommission. 2 August 2000.

DOUGLAS, D., 1998. “Will the Voters Restructure California’sLandmark Electricity Law ?”. PMA On Line Magazine. July 1998.

Energy Information Administration, 1999. “Status of State ElectricUtility Deregulation Activity”. US Department of Energy Web Page. 1September 1999.

Electricity Supply Association of Australia, 1999. “ElectricityAustralia 1999”. Sydney.Australia.

EU Commission, 1999a.“Second Harmonisation Report”. DG XVIIweb page. 15 August 1999.

EU Commission, 1999b.“Implementation of Electricity Directive byMember States”. DG XVII web page. 15 August 1999.

EU Commission, 2000.“Recent progress with building the internalelectricity market”. Brussels 16.5.2000. COM(2000) 297 final.

European Transmission System Operators, 1999. “InternationalExchanges of Electricity : Draft Rules Proposed By The EuropeanTransmission System Operators”. DG XVII Web Page. 15 August1999.

144

bibliography

FORD, A., 1999. “Cycles in Competitive Electricity Markets”.Energy Policy,Vol. 27, pp. 637-658.

FORD, A., 2000. “Understanding the Power Plant ConstructionCycle”. Public Utilities Fortnightly. 15 July 2000.

FRANKENA, M. and OWEN, B., 1994. “Electric Utility Mergers”.Praeger Publishers.

GRAY, P., 1997.“Colombia’s Mixed Ownership Model for PrivatizingInfrastructure”. The Private Sector in Infrastructure, 17-20.The WorldBank.

GREEN, R. and NEWBERY, D., 1992. “Competition in the BritishElectricity Spot Market”. Journal of Political Economy,Vol. 100, No. 5.

GREEN, R., 1996.“Increasing Competition in the British ElectricityMarket”. Journal of Industrial Economics, XLIV, 205-216.

GREEN, R., 1998. “Electricity Transmission Pricing : How much does itcost to get it wrong ?” Unpublished Working Paper. Department ofApplied Economics. Cambridge University.

GREEN, R. and McDANIEL, T., 1998. “Competition in ElectricitySupply : will 1998 be worth it ?” University of California EnergyInstitute Working Paper. PWP-048.

GUTH, L., 1998. “Market Power :An Exploration (Overview)”. TheElectricity Journal, Vol 11, No. 6.

HAUBRICH, H. and FRITZ, W., 1999. “Cross-Border ElectricityTransmission Tariffs”. Report to the EU Commission. DG TRENwebsite 15 August 2000.

HOGAN, W., 1999. “Electric Transmission Adequacy and MarketInstitutions”. Paper presented to the NARUC Summer Meetings,Committee on Electricity, on 20 July 1999. San Francisco.

HOGAN, W., 2000. “Transmission Rights and Market Power onElectric Power Networks : Comments”. Presentation at theResearch Conference on Electricity Industry Restructuring, 2000.University of California Energy Institute.

145

bibliography

HYMAN, L. and LLIC, M., 1997.“Scarce Resource, Real Business orThreat to Profitability ?”. Public Utilities Fortnightly. October 1997.

IEA, 1995. “Inter-system Competition and Trade in Electricity”.Paper presented to the Standing Group on Long-Term Co-operation. November 1995.

IEA, 1998. Electricity Market Reform.

JOSKOW, P. and SCHMALENSEE, R., 1983. Markets for Power. TheMIT Press.

JOSKOW, P., 1987.“Productivity Growth and Technological Changein the Generation of Electricity”. The Energy Journal, 8(1), 17-38.

JOSKOW, P., 1997. “Restructuring, Competition and RegulatoryReform in the US Electricity Sector”. Journal of EconomicPerspectives,. Vol 11, No. 3, 119-138.

JOSKOW, P. and TIROLE, J., 2000. “Transmission Rights and MarketPower on Electric Power Networks” Presentation at the ResearchConference on Electricity Industry Restructuring, 2000). Universityof California Energy Institute.

KING, J., 1997. “Metering in Real Time”. Public Utilities Fortnightly.September 1997.

KOOPS, R.J., 1999.“Whether US climate policy should be based ontaxes or permits”. Oxford Energy Forum.August 1999.

LAFFONT, J.J. and TIROLE, J., 1996.“Creating Competition throughInterconnection”. Journal of Regulatory Economics. 10 :227-256.

LEE, B., 1995. “Separability Test for the ESI”. Journal of AppliedEconometrics. 10 : 49-60.

LITTLECHILD, S.C., 1998. “Contribution to the EuropeanElectricity Regulation Forum”. 6 February 1998, Florence, Italy.

Ministry of Development, New Zealand, 2000. “Inquiry into theElectricity Industry”. June 2000.

146

bibliography

MOORE, I. and ANDERSON, J., 1997. “Introduction to the NewCalifornia Electricity Market”. Cal-PX web page, 14 May 1999.

National Electricity Code Administrator, 1999. “CapacityMechanisms in the National Electricity Market”. NationalElectricity Code Administrator (Australia).

NEWBERY, D., 1995. “Power Markets and Market Power”. TheEnergy Journal, 16(3) : 39-66.

NEWBERY, D. and POLLITT, M., 1997. “The Restructuring andPrivatization of the UK Electricity Supply : Was it worth it ?” ThePrivate Sector in Infrastructure, 77-80.The World Bank.

OECD/IEA, 1998.“Regulatory Reform in the US”.

OECD/IEA, 1999.“Regulatory Reform in Japan”.

Office of Electricity Regulation, 1996.“The Competitive ElectricityMarket from 1998 : Price Restraints”. Birmingham.

Office of electricity regulation, 1998.“Review of Electricity TradingArrangements : Proposals”. Birmingham.

Office of Gas and Electricity Markets, 2000. “Introduction of themarket abuse condition into the licences of certain generators.Ofgem’s second submission to the Competition Commission”.London.

OREN, S., 1999. “Ensuring Generation Adequacy in a CompetitiveElectricity Market”. Paper presented to the Workshop on Long-termGuarantee of Supply. Comision Nacional de la Energia. Madrid, 7June 1999.

O’ROURKE, 2000. “Redundant Restructuring : How the DualRetailer Model Makes Electric Markets Too Complex”. PublicUtilities Fortnightly, 1 July 2000.

PARKER, L. and ABEL, A., 1998. “Electricity : the Road TowardRestructuring”. CRS Issue Brief for Congress. National Institute forthe Environment.Washington D.C.

147

bibliography

PEREZ-ARRIAGA, J.I., 1998. “Visión global del cambio de regulación”.CSEN Working Paper DT 003/98. Madrid.

PEREZ ARRIAGA, J., DOLADER, J. and MAQUEDA, L., 1999.“Distribution Regulation In Competitive Environments :Investment, Pricing and Access”. CSEN Working Paper 001/99.Madrid.

PFEINBERGER, J., 1998.“In What Shape is Your ISO ?”The ElectricityJournal. July 1998.

PJM Interconnection, 1999.“Manual for Fixed Transmission Rights”.

Putnam, Hayes and Bartlett Consulting, 1997. “Summary ofInternational Transmission Pricing Approaches”. Paper presented toComision Nacional de la Energia. Madrid.

RADFORD,W., 1999.“20-odd Mergers in Search of a Policy”. PublicUtilities Fortnightly. 15 January 1999.

READ, E.G., 1998. “Electricity Sector Reform in New Zealand :Lessons from the Last Decade” Pacific and Asian Journal of Energy, 7(2) : 175-191.

RIVERA-BROOKS, N., 1999. “Power Rates Slow to Cool”. LosAngeles Times, 29 August 1999.

ROSENZWEIG, M. and VOLL, S., 1998. “Sequencing Power SectorPrivatization : is Reform Precondition or Result ?” Pacific and AsianJournal of Energy, 7 (2) : 119-132.

RUFF, L., 1999.“Capacity Payments, Stranded Costs and Security ofSupply”. Paper presented to the Workshop on Long-termGuarantee of Supply. Comision Nacional de la Energia. Madrid, 7June 1999.

SAUNDERS, B., 1999.“Capacity Payments in the England and WalesTrading Arrangements”. Paper presented to the Workshop onLong-term Guarantee of Supply. Comision Nacional de la Energia.Madrid, 7 June 1999.

148

bibliography

SCHWEPPE, F., CARAMANIS, M.,TABORS, R. and BOHN, R., 1988.Spot Pricing of Electricity. Norwell, Boston MA Kluwer.

SINGH, HARRY, 1999. “Capacity Payments, ISO Markets andGuarantee of Supply”. Paper presented to the Workshop on Long-term Guarantee of Supply. Comision Nacional de la Energia.Madrid, 7 June 1999.

SKYTTE, K., 1999. “The Regulating Power Market on the NordicPower Exchange NordPool : An Econometric Analysis”. EnergyEconomics 21, 295-308.

SMITH, W. and SHIN, B., 1995. “Funding Regulatory Agencies”.Economic Notes No.5.The World Bank. Country Dept. I.

SMITH, W., 1996. “Utility Regulators”. Paper presented to theInternational Forum for Utility Regulation. June 1996, Eynsham Hall(UK).

SURRATT, W., 1998. “Horizontal Market Power”. The ElectricityJournal, Vol. 11, No. 6.

THOMAS, S., 1999.“Has Privatisation Reduced the Price of Powerin Britain ?” November 1999. http ://www.unison.org.uk/.

VALLETTI, T. and ESTACHE, A., 1998. “The Theory of AccessPricing : An overview for Infrastructure Regulators”. Paperprepared for The World Bank Institute.

VAN VACTOR, 2000. “Comparing Electricity Markets in Californiaand PJM”. Public Utilities Fortnightly. 15 July 2000.

VISCUSI, K.,VERNON, J. and HARRINGTON, J., 1995. Economics ofRegulation and Antitrust. MIT Press.

WISER, R., GOLOVE,W. and PICKLE, S., 1998.“California’s ElectricMarket : What’s in it for the Consumer”. Public Utilities Fortnightly.August 1998.

149

bibliography

WOLAK, F.A., 1997. “Electricity Market Structure and Pricing : anInternational Comparison”. Paper presented to Energy ModellingForum 15, 11-12 September 1997, Stanford University.

WOLFRAM, C., 1999. “Measuring Duopoly Power in the BritishElectricity Spot Market.” American Economic Review, 89 : 805.

WOLFRAM, C., 2000.“Electricity Markets : Should the Rest of theWorld Adopt the United Kingdom’s Reforms ?” Regulation, 22 : 48.

World Trade Organisation, 1998. “Energy Services”. BackgroundNote by the Secretariat, 9 September 1998.

ZACCOUR, G., 1998. Deregulation of Electric Utilities. Boston :Kluwer.

150

bibliography

GLOSSARY71

Access Charge: A charge levied on a power supplied, or itscustomer, for access to a utility’s transmission or distributionsystem. It is a charge for the right to send electricity over another’swires.

Ancillary Services: Services that are required for the securityand stability of the transmission system, such as reactive power, hotstandby and frequency control.

Base Load:The minimum load over a given time period (e.g. a dayor a year).

Bilateral Contract: A direct contract between a powerproducer and user or broker outside of a centralised power poolor POOLCO.

Broker: An entity acting as an agent for others in negotiatingcontracts, purchases, or sales of electrical energy or serviceswithout owning any transmission or generation facilities. Brokersdo not own the power transacted.

Bulk Power Market: Purchases and sales of electricity amongelectricity companies. Often this term is used interchangeably withwholesale power market.

Capacity Payment: A payment to generators proportional tothe generating capacity they make available.

Captive Customer: Those customers who do not have anoption to go to another supplier to buy electricity.

CHP: Combined Heat and Power, or Cogeneration.

Cogeneration: A process of producing simultaneously electricand thermal (heat) energy from one fuel source.

151

glossary

71. Many of the definitions in this glossary have been taken from other sources including the glossarieselaborated by NARUC, Offer, California ISO, Entergy, Edison Electric Institute, US DOE and Australia’sParliamentary Library.

Combined Cycle: An electric generating technology in whichelectricity is produced from otherwise lost waste heat exiting fromone or more gas (combustion) turbines.

Congestion: See transmission congestion.

Contract for Differences: A financial contract that enablescustomers to purchase power at a fixed price.

Contract Path: The most direct physical transmission tiebetween two interconnected entities. In some electricitytransactions, the transfer of power is presumed to take placeacross the “contract path”.

Corporatisation: Subjecting government business enterprises tothe principles of corporate law, often accompanied by a range ofother initiatives, such as providing greater management autonomyand clear commercial objectives.

Cost of Service Regulation: A form of regulation where theprofit of a regulated company is determined by its actual costsregardless of performance. Tariffs are set to allow the firm torecover its costs, including the cost of capital.

Countertrade: See Redispatching.

Demand-Side Management: See DSM.

Deregulation: The elimination of regulation from a previouslyregulated industry or sector of an industry.

Direct Access: The ability of a customer to purchase electricitydirectly from the wholesale market rather than through a localdistribution utility. Customers with direct access are also said to be“eligible”.

Distributed Generation: This refers to generation units locatednear a consumption point so that power can be delivered withoutmaking use of the transmission grid.

Distribution: The process of transferring electricity from thetransmission grid to final users.

152

glossary ===

Distribution Company (Disco): The regulated electriccompany that constructs and maintains the distribution wiresconnecting the transmission grid to the final customer.

Divestiture: Removing one function from others by selling(spinning-off), or in some other way changing the ownership, of theassets related to that function. For instance, spinning-off generationassets so they are no longer owned by the shareholders who ownthe transmission and distribution assets.

DSM (Demand-Side Management): Planning, implementation,and evaluation of utility-sponsored programmes to influence theamount or timing of customers’ energy use.

Economic Efficiency: A term that refers to the optimalproduction and consumption of goods and services.This generallyoccurs when prices of products and services reflect their marginalcosts.

Economies of Scale: An industry (or technology) exhibitseconomies of scale if marginal cost decreases with output.

Eligible Customer: Those customers who have the option to goto another supplier to buy electricity.

Embedded Cost: See sunk Cost.

End User Supply: See retail supply.

Excess Capacity: Volume or capacity over and above that whichis needed to meet peak (planned or expected) demand.

Forwards: A “forward” is a commodity bought and sold fordelivery at some specific time in the future. It is differentiated froma futures contract in that a forward contract is a customised, non-exchange traded, and non-regulated hedging mechanism.

Fossil Fuel: Any naturally occurring organic fuel, such aspetroleum, coal, and natural gas.

Futures Market: An arrangement through a standardisedcontract for the delivery of a commodity at a future time and at a

153

glossary

price specified at the time of purchase. The price is based on anauction or market basis.

Generation: The process of converting primary energy (e.g. coal,gas, oil, stored water, wind and solar) into electricity.

Generation Company (Genco): An entity that operates andmaintains generating plants.

Generation Dispatch and Control: Aggregating anddispatching (sending off to some location) generation from variousgenerating facilities, providing backup and reliability services.

Grid: A network for the transmission of electricity throughout thestate or country typically consisting of several interconnectedlines.Also, the high voltage transmission network.

Incentive Regulation: A form of regulation where the profit ofa regulated company is determined by its actual performancerelative to some pre-established standards of performance forservice quality and cost effectiveness.

Independent System Operator (ISO): A neutral operatorresponsible for maintaining instantaneous balance of the gridsystem.The ISO performs his function by controlling the dispatchof flexible plants to ensure that loads match resources available tothe system.

Interconnector: High voltage transmission lines linking states orcountries.

Load: The amount of electricity being used or demanded at onetime by a circuit or system.

Load Profiling: The study of the consumption habits ofconsumers to estimate the amount of power they use at varioustimes of the day and for which they are billed. Load profiling is analternative to precise metering.

Locational Market Clearing Price: The price at which supplyequals demand at a specified location. All demand which is

154

glossary ===

prepared to pay at least this price at the specified location has beensatisfied. All supply which is prepared to operate at or below thisprice in the specified location has been purchased.

Losses: Electric energy transformed into heat and therefore lostduring transportation.

Marginal Cost: The cost of providing the next (marginal)kilowatt-hour of electricity, irrespective of sunk costs.

Market Power: A company’s use of its position in a market toraise prices above competitive levels.

Merger: The union of two or more commercial interests orcorporations.

Merit Order: Ranking in order of which generation plant shouldbe used, based on ascending order of price together with amountof energy that will be generated.

Monopoly: The only seller with control over market sales.

Natural Monopoly: A situation where one firm can produce agiven level of output at a lower total cost than can any combinationof multiple firms. Natural monopolies occur in industries whichexhibit decreasing average long-run costs due to size (economiesof scale).

Nodal Pricing: Locational pricing of energy and transmissionservices (See Locational Market Clearing Price).

Oligopoly: A few sellers who exert market control over prices.

Options: A contractual agreement that gives the holder the rightto buy (call option) or sell (put option) a fixed quantity of asecurity or commodity (e.g., a commodity or commodity futurescontract), at a fixed price, within a specified period of time. Optionsmay be standardised, exchange-traded, and government regulated,or over-the-counter customised and non-regulated.

155

glossary

Pancaking: Charging two or more access charges to electricitytransactions that make use of two or more transmission systems.Pancaking discourages intersystem trade and does not generallyreflect transmission costs.

Peak Load (or Peak Demand): The electric load thatcorresponds to a maximum level of electric demand in a specifiedtime period.

Performance-Based Regulation (PBR): See IncentiveRegulation.

Pool: A short-term market for electricity where sellers bid intothe pool the advance prices of quantities of electricity, andgenerators are dispatched to meet the demand.A pool comprisesthe functions of a power exchange and a system operator. Thesefunctions can be performed by a single entity or, alternatively, canbe unbundled.

Poolco: An entity that operates a pool.

Postage Stamp Tariff: In transmission, a price for transmissionservices that does not depend on location. Paying the tariff givesthe right to inject energy at any node of the grid and take it at anyother node.

Power Exchange (PX): An independent entity responsible forconducting an auction for generators seeking to sell energy and forloads which are not otherwise being served by bilateral contracts.The Power Exchange is generally responsible for schedulinggeneration, determining market clearing prices, and for settlementand billing.

Price Cap: A price determined by the regulator that cannot beexceeded by the regulated company. It serves to implementincentive regulation.

Rate of Return Regulation: See Cost of Service Regulation.

Redispatching: Management of transmission congestion directlyby the system operator.This consists of requiring some generating

156

glossary ===

units located in the zone where demand exceeds supply toincrease output while reducing the output of some generators inthe zone where supply exceeds demand.

Reliability: Electricity system reliability has two components,adequacy and security. Adequacy is the ability of the system tosupply the aggregate electrical demand and energy requirements ofthe customers at all times, taking into account scheduled andunscheduled outages of system facilities. Security is the ability ofthe system to withstand sudden disturbances such as electric shortcircuits or unanticipated loss of system facilities.

Renewable Resources: These are energy resources that arenaturally replenishable, but flow-limited. They are virtuallyinexhaustible in duration but limited in the amount of energy thatis available per unit of time. Renewable energy resources include:biomass, hydro, geothermal, solar and wind. In the future they couldalso include the use of ocean thermal, wave, and tidal actiontechnologies.

Retail Competition: A system under which more than oneelectricity provider can sell to retail customers, and retailcustomers can buy from more than one provider.

Retail Market: A market in which electricity and other energyservices are sold to the end-user.

Retail Supply: The business of purchasing electricity at bulksupply points and selling it to retail customers.

“Split the Market” Approach: Managing transmissioncongestion by means of locational or nodal pricing.

Spot Market: A market for the immediate delivery of acommodity. Short-term electricity markets are sometimes(incorrectly) denominated “spot electricity markets” even ifdelivery does not take place immediately.

157

glossary

Stranded Assets: Unamortised ESI assets that will not generateenough revenue in a deregulated market to allow for full costrecovery.

Sunk Cost: A cost that has already been incurred, and thereforecannot be avoided or changed by any strategy going forward.

System (Electricity): Physically connected generation,transmission, and distribution facilities operated as an integratedunit under one central management or operating supervision.

System Operation: The process of maintaining instantaneousbalance of an electricity system.

Tariff: A regulated price and any additional regulated serviceconditions linked to it.

Third Party Access: Right to equal (non-discriminatory) accessto transmission services.

Time-of-Use (TOU) Rates: The pricing of electricity based onthe estimated cost of electricity during a particular time block.Time-of-use rates are usually divided into three or four timeblocks per twenty-four hour period (on-peak, mid-peak, off-peak,and sometimes super off-peak) and by seasons of the year(summer and winter). Real-time pricing differs from TOU rates inthat it is based on actual (as opposed to forecasted) prices whichmay fluctuate many times a day and are weather-sensitive, ratherthan varying with a fixed schedule.

Transco: An independent transmission company that is engagedsolely in the bulk transmission of electricity, owns transmissionassets and manages system operation. It can be for-profit or not.

Transmission:The process of transporting electric energy in bulkfrom a source of supply to other parts of the system or to othersystems.

Transmission Congestion: The condition when marketparticipants seek to dispatch in a pattern which would result in

158

glossary ===

power flows that cannot be physically accommodated by thesystem. Although a system will not normally be operated in anoverloaded condition, it may, nevertheless, be described ascongested based on requested/desired schedules.

Transmission Congestion Contract (TCC): A financialinstrument that provides a hedge against congestion pricedifferences between zones or nodes of the grid.

Transmission System Operator (TSO): See SystemOperator.

Unbundling: Disaggregating a service into its basic componentsand offering each component separately for sale with separaterates for each component. For example, generation, transmissionand distribution could be unbundled and offered as discreteservices.

Utility: A regulated company that is the monopoly supplier ofsome service. Today this term is also applied to companies thatoperate in former utility industries.

Vertical Integration: An arrangement whereby the samecompany owns all the different aspects of making, selling, anddelivering a product or service. In the electricity industry, it refersto the historically common arrangement whereby a utility wouldown its own generating plants, transmission system, anddistribution lines to provide all aspects of electricity service.

Wheeling: Transmission of electricity by an entity that does notown, or directly use, the power it is transmitting. Wholesalewheeling is used to indicate bulk transactions in the wholesalemarket, whereas retail wheeling allows power producers directaccess to retail customers.This term is often used colloquially tomean transmission.

Wholesale Competition: A system whereby a distributor ofpower would have the option to buy his power from a variety of

159

glossary

power producers, and the power producers would be able tocompete to sell their power to a variety of distribution companies.

Wholesale Power Market: The purchase and sale of electricityfrom generators to resellers (who sell to retail customers) alongwith the ancillary services needed to maintain reliability and powerquality at the transmission level.

160

glossary ===

DELIVERY DETAILS

Name Organisation

Address

Country Postcode

Telephone Fax

PAYMENT DETAILS

I enclose a cheque payable to IEA Publications for the sum of US$ or FF

Please debit my credit card (tick choice). Access/Mastercard Diners VISA AMEX

Card no:

Expiry date: Signature:

I would like to order the following publications

OECD CENTRES

Please send your order

by mail, fax, or by e-mail

to your nearest

OECD Centre

OECD TOKYO CENTRE

Landic Akasaka Building

2-3-4 Akasaka, Minato-ku

Tokyo 107-0052, Japan

Tel: (+81-3) 3586 2016

Fax: (+81-3) 3584 7929

E-mail: [email protected]

Internet: www.oecdtokyo.org

OECD WASHINGTON CENTER

2001 L Street NW, Suite 650

Washington, D.C., 20036-4922, US

Tel: (+1-202) 785-6323

Toll-free number for orders:

(+1-800) 456-6323

Fax: (+1-202) 785-0350


Internet: www.oecdwash.org

OECD BONN OFFICE

c/o DVG mbh (OECD)

Birkenmaarstrasse 8

D-53340 Meckenheim, Germany

Tel: (+49-2225) 926 166

Fax: (+49-2225) 926 169


Internet: www.oecd.org/bonn

OECD MEXICO CENTRE

Edificio INFOTEC

Av. Presidente Mazarik 526

Colonia: Polanco

C.P. 11560 - Mexico D.F.

Tel: (+52-5) 280 12 09

Fax: (+52-5) 280 04 80


Internet: www.rtn.net.mx/ocde

Order FormOrder Form

PUBLICATIONS ISBN QTY PRICE* TOTAL

Competition in Electricity Market 92-64-18559-3 $75

Regulatory Reform: European Gas 92-64-18558-5 $75

Electricity Market Reform: Power Generation Costs and Investment 92-64-16961-X $50

Electricity Market Reform - An IEA Handbook 92-64-16187-2 $50

World Energy Outlook 2000 92-64-18513-5 $150

Electric Power Technology - Opportunities and Challenges of Competition 92-64-17133-9 $40

Energy Policies of IEA Countries – 2000 Review (Compendium) 92-64-18565-8 $120

Business as Usual and Nuclear Power 92-64-17175-4 $25

TOTAL

*Postage and packing fees will be added to each order.

IEA PUBLICATIONS, 9, rue de la Fédération, 75739 PARIS Cedex 15Photo PIX - Printed in France by Jouve(61 00 27 1) ISBN 92-64-185593 2001

Imprimé en France – JOUVE, 18 rue Saint-Denis, 75001 Paris – N° 288882D - Dépôt légal : Février 2001

Competition in Electricity Markets - Regulation Body of

Documents

Transcript of Competition in Electricity Markets - Regulation Body of